ASUS Maximus II Formula - new P45 king?

[kamieldehond]

Increasing the TRFC memory timing is supposed to help a lot with 4 sticks, have you tried vaues over 50?

tried it and it worked a little. got further trough memtest v2.11 but it craps out at test 5: random moves.

only running my d9gmh at 2.1 volts as im a little scared to destroy them (as i did with my previous pair of d9gmh at only 2.2 volts :down:) but i do have a memory cooler installed so that will help :)

[In-Fluence]

It now transpires that going back to the 1307 bios has made this board behave how I believe it should! I've no idea why, but my Quad and reapers seem to prefer this one to 1901 & 2104. I've now got it prime blending for more than an hour at 410x9 with the same CPU VTT and RAM voltages I had on the P5B, (1.45/1.2/2.1) and with some half-decent memory speeds too. I'm getting hangs as I would have expected with given voltages & speeds and everything is making a lot more sense to me now. Once the skews and GTLs are tuned I hope to drop some temps & raise the FSB a little - damn this hot weather though, it's really holding me back at the moment!

For the record, an afterthought occurred about the whole matter with possible bad bios flashes from USB as a result of having my printer plugged in at the same time (which wasn't when I flashed to 1307). This may have caused an issue which I intend to investigate further as in my hastiness it never really crossed my mind. If it turns out that this was the cause and the newer bios behaves in the same way as 1307 does now, I'll stick with 2104 as I got less hangs on reboot. (I'll also feel a bit stupid too) :D

What kind of NB voltages are you running Kamieldehond? Perhaps fill out an entire setings template with the most stable ones you got so far....

[humeyboy]

I seriously doubt having any other USB devices plugged in and powered up will cause any issues flashing the Bios.

1307 is said to be better, some claim its the best, 1901 is said to be good and the 2 newer ones (1 was pulled) are said to be crap.

[In-Fluence]

That's exactly what I thought, but I'm wondering what might actually be the issue. I've left 2104 priming on the same settings as I did with 1307, and it had done 10 mins before I left it. When I return from work, I shall see if this may have affected it, but I would expect to be returning to a restarted/hung computer. If not then perhaps I could be right as I've done nothing different other than that.

I booted from the USB stick as well as flashed from it, and the printer is one of those lexmark all in one scanner/copier things (which I hate, but it was free), and it uses a weird driver. It wouldn't suprise me if it used some kind of sub-standard hardware that could possibly interfere with it, but then again, I could have just had a few unlucky flashes!

[Grnfinger]

I seriously doubt having any other USB devices plugged in and powered up will cause any issues flashing the Bios.

1307 is said to be better, some claim its the best, 1901 is said to be good and the 2 newer ones (1 was pulled) are said to be crap.

What bios are you running?
I have 1802 on bios 1 and 1901 on bios 2 I may revert back to 1307 to test some ram clocks...seems on 1901 my 8800's wont clock as high

[kamieldehond]

Ai Overclock Tuner: Manual
OC From CPU Level Up: auto
Ratio CMOS Setting: 9.5
FSB Frequency: 450
CPU Clock Skew: Auto
NB Clock Skew: Auto
FSB Strap to North Bridge: auto
DRAM Frequency: 1128mhz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual
1st Info: 5-5-5-15
2nd Info: Auto
3rd Info: Auto



DRAM Static Read Control: Auto
DRAM Read Training: Auto
MEM. OC Charger: Auto
Ai Clock Twister: Auto
Ai Transaction Booster : Auto


PCIE Frequency : 100

CPU Voltage: 1.375
CPU PLL Voltage:1.553
FSB Termination Voltage:1.24575
DRAM Voltage: 2.10475
North Bridge Voltage:1.32525
South Bridge 1.5 Voltage: Auto
South Bridge 1.1 Voltage: Auto

CPU GTL Reference (0): Auto
CPU GTL Reference (1): -40mv
CPU GTL Reference (2): Auto
CPU GTL Reference (3): -40mv
NB GTL Reference: -60mv
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto

Load-Line Calibratie: Enabled
both Spread Spectrum Disabled

but ill probably will sell the board as i have a week spot for DFI and will buy a UT P45-T2RS as they are pretty cheap new

[humeyboy]

What bios are you running?
I have 1802 on bios 1 and 1901 on bios 2 I may revert back to 1307 to test some ram clocks...seems on 1901 my 8800's wont clock as high

2104 :shrug:

[In-Fluence]

Hello again - Here's my suggestion going by my limited experience of this board -

Ai Overclock Tuner: Manual
OC From CPU Level Up: auto
Ratio CMOS Setting: 9.5
FSB Frequency: 450
CPU Clock Skew: Auto
NB Clock Skew: Auto
FSB Strap to North Bridge: auto
DRAM Frequency: 1128mhz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual
1st Info: 5-5-5-15 - Fill out the rest rather than leaving on auto and set TRFC to 55 or maybe a lot more
2nd Info: Auto
3rd Info: Auto (and the same for these two)


Most of the following settings are pretty important in order to rule out memory issues:
DRAM Static Read Control: Disable
DRAM Read Training: Disable
MEM. OC Charger: Enable
Ai Clock Twister: Moderate
Ai Transaction Booster : Manual
Performance level: 10 (maybe 11 or more too)

PCIE Frequency : 100

CPU Voltage: 1.375
CPU PLL Voltage:1.553
FSB Termination Voltage:1.24575
DRAM Voltage: 2.10475 - [try 2.11 too, this should be ok for 24/7]
North Bridge Voltage:1.32525 - Perhaps nudge this up to 1.35-1.45
South Bridge 1.5 Voltage: Auto
South Bridge 1.1 Voltage: Auto

CPU GTL Reference (0): Auto
CPU GTL Reference (1): -40mv
CPU GTL Reference (2): Auto
CPU GTL Reference (3): -40mv
NB GTL Reference: -60mv
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto Test these as a pair & see if values between +/- 25 help
North Bridge DDR Reference: Auto Test this too but with +/-50 as your limit

Load-Line Calibratie: Enabled
both Spread Spectrum Disabled

As you can probably tell, I'm making rapid progress with this board now, and hope to have some half decent, fully stable settings soon :)

[Kayso]

Latest bios:

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 475
CPU Clock Skew : Auto
North Bridge Clock Skew : Auto
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-1141
Dram Clock Skew CH1 A1 : Auto
DRAM Clock Skew CH1 A2 : Auto
Dram Clock Skew CH1 B1 : Auto
Dram Clock Skew CH1 B2 : Auto

DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time : 55
Write Recovery Time : 6
Read to Precharge Time : 3

Read to Write Delay (S/D) : Auto
Write to Read Delay (S) : Auto
Write to Read Delay (D) : Auto
Read to Read Delay (S) : Auto
Read to Read Delay (D) : Auto
Write to Write Delay (S) : Auto
Write to Write Delay (D) : Auto

Write to PRE Delay : Auto
Read to PRE Delay : Auto
PRE to PRE Delay : Auto
ALL PRE to ACT Delay : Auto
ALL PRE to REF Delay : Auto

DRAM Static Read Control: Enabled
Dram Read Training : Enabled
MEM OC Charger : Disabled
Ai Clock Twister : Stronger
Transaction Booster : Manual

Common Performance Level [8]

Pull-In of CHA PH1 Disabled
Pull-In of CHA PH2 Disabled
Pull-In of CHA PH3 Disabled
Pull-In of CHA PH4 Disabled
Pull-In of CHA PH5 Disabled
Pull-In of CHB PH1 Disabled
Pull-In of CHB PH2 Disabled
Pull-In of CHB PH3 Disabled
Pull-In of CHB PH4 Disabled
Pull-In of CHB PH5 Disabled

PCIE Frequency : 101

CPU Voltage : 1.4375
CPU PLL Voltage : 1.55
FSBT : 1.325
DRAM Voltage : 1.96
North Bridge Voltage : 1.45
South Bridge Voltage 1.5 : 1.55
South Bridge Voltage 1.1 : 1.15

CPU GTL Reference 0 : Auto
CPU GTL Reference 1 : Auto
CPU GTL Reference 2 : Auto
CPU GTL Reference 3 : Auto
North Bridge GTL Reference : +60mv
DDR2 Channel A REF Voltage : AUTO
DDR2 Channel B REF Voltage : AUTO
North Bridge DDR Reference : AUTO

Load Line Calabration : Enabled
CPU Sread Spectrum : Disabled
PCIE Spread Spectrum : Disabled


Those are mine 24/7 settings for now. Anything above 475 FSB fails Prime, no matter on what voltages or GTL's neither Skew's... And it is not my CPU, because I have the same problems on 6*480 as an example, on 960 Mhz memory. So it is the motherboard or NB.
But I'm happy with those 475 after all :)

[kamieldehond]

ok played around with the bord and the memory a little

Ai Overclock Tuner: Manual
OC From CPU Level Up: auto
Ratio CMOS Setting: 9.5
FSB Frequency: 450
CPU Clock Skew: Auto
NB Clock Skew: Auto
FSB Strap to North Bridge: Auto
DRAM Frequency: 1200mhz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual
1st Info: 5-5-5-18-5-60-5-3
2nd Info: 8-3-5-4-6-4-7
3rd Info: 13-5-1-5-5



DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Enabled
Ai Clock Twister: Moderate

Preformance lvl: 10


PCIE Frequency : 100

CPU Voltage: 1.375
CPU PLL Voltage:1.553
FSB Termination Voltage:1.24575
DRAM Voltage: 2.21075
North Bridge Voltage:1.40475
South Bridge 1.5 Voltage: Auto
South Bridge 1.1 Voltage: Auto

CPU GTL Reference (0): Auto
CPU GTL Reference (1): -40mv
CPU GTL Reference (2): Auto
CPU GTL Reference (3): -40mv
NB GTL Reference: -60mv
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto

Load-Line Calibratie: Enabled
both Spread Spectrum Disabled

This passes memtest running 4xd9gmh :up:

I must admit i like the board :up:

[Grnfinger]

Those are mine 24/7 settings for now. Anything above 475 FSB fails Prime, no matter on what voltages or GTL's neither Skew's... And it is not my CPU, because I have the same problems on 6*480 as an example, on 960 Mhz memory. So it is the motherboard or NB.
But I'm happy with those 475 after all :)

are you running GSkill ram still? your sig list GSkill thats why I ask...if you are running GSkill what version and kit?

[Kayso]

I don't know what version, but my memory model is [PI] F2-8800CL5D-4GBPI: http://www.gskill.com/products.php?index=165

Why are you asking :)

[Grnfinger]

I have the same kit and they used to run 1145/1150MHz now they seem to top out at 1130MHz wondering if its the bios 1901 I'm on

[LennyRhys]

I've been priming at 4GHz, 1.21625v in BIOS, and after 30mins the PC just restarts. I'm amazed the CPU is stable at 1.216v under load, but something is making it give up...

I'm using BIOS 907... haven't flashed since I got the board but I will if there is a better BIOS. What's the best BIOS for Q9650, in people's experience?

[Grnfinger]

I've been priming at 4GHz, 1.21625v in BIOS, and after 30mins the PC just restarts. I'm amazed the CPU is stable at 1.216v under load, but something is making it give up...

I'm using BIOS 907... haven't flashed since I got the board but I will if there is a better BIOS. What's the best BIOS for Q9650, in people's experience?

ewww 907 was a dog get off that bios...1307 1802 and 1901 were very good, I have not tested anything recent
If prime is restarting I would bump your vCore up a few notches, I need 1.325 for 4.2GHz

[Badmuts]

Hi Guys,

I've some problem with my board (I think) Now I'm DICE benching with my Maximus II Forumla and QX6700 the Vcore voltage won't boot above 1.7000v (BIOS) but anything below works perfect. (like 1.69v)

Sorry for my bad English but I hope you understand and you can help me.

[Grnfinger]

What bios version?

[Badmuts]

Just updated to 2104 and still the same problem :( (Updated from: 1901)

[LennyRhys]

ewww 907 was a dog get off that bios...1307 1802 and 1901 were very good, I have not tested anything recent
If prime is restarting I would bump your vCore up a few notches, I need 1.325 for 4.2GHz

Yeah I worried that's what it was... bumped vcore from 1.225 to 1.2375 in BIOS, reads 1.240 in CPU-Z under load, blend stable for 1 hour so it's definitely an improvement.

I'll flash to a newer BIOS and see if I get any improvement.

[Kayso]

I have the same kit and they used to run 1145/1150MHz now they seem to top out at 1130MHz wondering if its the bios 1901 I'm on

Mine is almost the same. 1145-1150 are maximum. But memory is not the problem. With GSkill 1200 mhz I have the same issues. Over 475 Fsb - impossible to make it stable, no matter voltages and skew's... LOLL I boot in windows at 500 FSB, I can make screens, do some benching, but when I start Prime Blend... errors on the same second.

[LennyRhys]

Something I forgot to mention - since replacing the motherboard TIM with my own Zalman STG1 paste, the NB and SB temps have dropped by about 10C. :eek:

[Stormad]

Hi guys, i've finally been accepted in your forum so now I can thank you all for your generous contributions in my overclocking adventure with the MaxFormII, especially the simps guide to finetune all the settings and it also took me a better processor for getting higher clocks because my first processor was a E8400 C0 and the max stable was 3880Mhz even at 1.47V. And with my new E8500 E0 I'm running 4302Mhz 24/7 at 1.38V it will even go 4555Mhz at around 1.42V. But I'm very happy with the 4302Mhz. I'll try to post a screen of everest too.
http://img198.imageshack.us/img198/8135/superpi32m658235.th.jpg (http://img198.imageshack.us/i/superpi32m658235.jpg/)

[In-Fluence]

Grrrr! I'm getting really frustrated with this board now. Seems it was prime blending for 6 1/2 hours last night, but woke up to it restarted again. There was more than enough Vcore for 456x8 (1.4625), northbridge at 1.4, vtt at 1.32 (have tried higher but seems to fail quicker), and PLL at 1.59.. Ram was well within limits, and my CPU GTLs are tuned. CPU temps were 30° idle and 65° load and it happily blended at the same volts and higher temps on the p5b-dlx (which has no GTL tuning).... This was with 2104 so I think I'm gonna try moving back to 1901 or 1307. Other than that - any hints?

That's a nice o/c you got there Stormad - I'm might borrow an E8500 this weekend which I got to 548 FSB stable on my P5b - if the board doesnt handle that too well either I'm gonna be very dissappointed and might even move back to my P5b dlx!

[WhiteFireDragon]

Something I forgot to mention - since replacing the motherboard TIM with my own Zalman STG1 paste, the NB and SB temps have dropped by about 10C. :eek:

that's quite a nice drop in temps just for TIM replacement. what were your temps now? and do you think you can do a quick screen shot of showing all the PC probe values?


Hi guys, i've finally been accepted in your forum so now I can thank you all for your generous contributions in my overclocking adventure with the MaxFormII, especially the simps guide to finetune all the settings and it also took me a better processor for getting higher clocks because my first processor was a E8400 C0 and the max stable was 3880Mhz even at 1.47V. And with my new E8500 E0 I'm running 4302Mhz 24/7 at 1.38V it will even go 4555Mhz at around 1.42V. But I'm very happy with the 4302Mhz. I'll try to post a screen of everest too.
http://img198.imageshack.us/img198/8135/superpi32m658235.th.jpg (http://img198.imageshack.us/i/superpi32m658235.jpg/)

:welcome:

glad the tuning gave you some pretty nice improvements :up:. most important for dual cores is nb gtl and cpu/nb skews. i only found that the cpu gtl mattered for quads, and did nothing for dual cores

[LennyRhys]

I'm currently trying to get 478FSB stable, using the 7.5 multi for now (3585 mhz). Will be working on it for a while I think...

[Grnfinger]

I'm currently trying to get 478FSB stable, using the 7.5 multi for now (3585 mhz). Will be working on it for a while I think...

looking forward to your results, keep us posted

[LennyRhys]

that's quite a nice drop in temps just for TIM replacement. what were your temps now? and do you think you can do a quick screen shot of showing all the PC probe values?

The stock TIM was shocking - almost as bad as what I found under my 680i NB heatsink - so I shouldn't really be surprised by the drop in temps. My chipset temps generally go to about 50C under load; before replacing the TIM, that's what they were at idle. :eek: I'll be posting my OC results soon because I've settled on speed and just need to tweak the voltages, so I'll put the SB and NB temps in. :)


looking forward to your results, keep us posted

I've been following your BIOS settings for 4.2 pretty closely. I think 478FSB will require a lot more time and effort, but more importantly the CPU gets too hot... I was priming before at 4300MHz and cores were all about 75C within a few minutes...and then core 0 failed. I'm probably just going to stick with 4.2 and see how low I can get my vcore.

Thanks for all the help. :up:

[LennyRhys]

Might need a little help... I've decided to go for 4.1GHz on the 8.5 multi - much higher memory bandwidth and a lower vcore.

So far I've managed to get prime blend running for about 10-15 minutes then the system just freezes, it doesn't error or reboot etc.

Here are my settings for 483x8.5 - vcore can easily come down but I need to get the FSB stable first...

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 8.5
FSB Frequency : 483
CPU Clock Skew : Delay 200ps
North Bridge Clock Skew : Delay 200ps
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-1160
Dram Clock Skew CH1 A1 : Normal
DRAM Clock Skew CH1 A2 : Normal
Dram Clock Skew CH1 B1 : Normal
Dram Clock Skew CH1 B2 : Normal


DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : Auto
Row Refresh Cycle Time : Auto
Write Recovery Time : Auto
Read to Precharge Time : Auto

Read to Write Delay (S/D) : Auto
Write to Read Delay (S) : Auto
Write to Read Delay (D) : Auto
Read to Read Delay (S) : Auto
Read to Read Delay (D) : Auto
Write to Write Delay (S) : Auto
Write to Write Delay (D) : Auto

Write to PRE Delay : Auto
Read to PRE Delay : Auto
PRE to PRE Delay : Auto
ALL PRE to ACT Delay : Auto
ALL PRE to REF Delay : Auto

DRAM Static Read Control: Enabled
Dram Read Training : Enabled
MEM OC Charger : Disabled
Ai Clock Twister : Moderate
Transaction Booster : Manual

Common Performance Level [8]

Pull-In of CHA PH1 Disabled
Pull-In of CHA PH2 Disabled
Pull-In of CHA PH3 Disabled
Pull-In of CHA PH4 Disabled
Pull-In of CHA PH5 Disabled
Pull-In of CHB PH1 Disabled
Pull-In of CHB PH2 Disabled
Pull-In of CHB PH3 Disabled
Pull-In of CHB PH4 Disabled
Pull-In of CHB PH5 Disabled

PCIE Frequency : 102

CPU Voltage : 1.3000
CPU PLL Voltage : 1.57
FSBT : 1.36
DRAM Voltage : 1.83
North Bridge Voltage : 1.45
South Bridge Voltage 1.5 : 1.55300
South Bridge Voltage 1.1 : 1.20600

CPU GTL Reference 0 : Auto
CPU GTL Reference 1 : Auto
CPU GTL Reference 2 : Auto
CPU GTL Reference 3 : Auto
North Bridge GTL Reference : +60mv
DDR2 Channel A REF Voltage : AUTO
DDR2 Channel B REF Voltage : AUTO
North Bridge DDR Reference : AUTO

Load Line Calabration : Enabled
CPU Sread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

[Kayso]

Might need a little help... I've decided to go for 4.1GHz on the 8.5 multi - much higher memory bandwidth and a lower vcore.

So far I've managed to get prime blend running for about 10-15 minutes then the system just freezes, it doesn't error or reboot etc.

Here are my settings for 483x8.5 - vcore can easily come down but I need to get the FSB stable first...



When your system freezes, mean that you are lacking vcore. Bump vcore 3 steps further and I think you will be okay.

But you have Q9650 with multi 9. Why the hell are you working with 8.5 multi ?!?! This motherboard goes with 9*475 without any problems with all skews and GTL's on auto. 9*475 = 4275 and with little tweaking you can do 479, which is over 4300 mhz. Whit this memory I'm pretty sure you can do 485 FSB without any problems.

[LennyRhys]

I tried your bios settings for 475 and the fsb is not stable on my board with these settings. As for vcore being the problem, my system is (at least) 1hr Prime stable at 4.2GHz with only 1.318v in bios so I didn't think I'd need anything like 1.3 vcore for 4.1GHz, but I'll bump it up and see what happens...

I'm working with the 8.5 multi because I want high memory bandwidth, I need too much vcore for 4.3GHz and it's far too hot. :yepp:

I'm also going to try tuning the clock skews and GTLs to see if I can prime for any longer - with CPU and NB skews at Delay 200ps I can prime for 15mins, but set to Auto the PC freezes within 1 minute

[CryptiK]

Lenny also search the thread if you haven't already done so for andressergio's settings, he was achieving some good results with his Q9650 before he changed boards.

[LennyRhys]

Yeah I've been trying andressergio's settings and they have been very helpful... in particular I'd say finding the right NB GTL has really marked the cusp of stability for my system...

+80mv - all cores fail in a few seconds
+70mv - unstable
+60mv - unstable
+50mv - perfect

However I think I've been over-complicating things - Kayso was right, it was my vcore holding me back :mad:

All these hours spent tweaking PLL, VTT etc and all I really needed to do was bump vcore up a couple of notches.

Surprisingly 4105mhz 483x8.5 needs exactly the same vcore as 4200mhz 467x9 - 1.32v - which is frustrating...

Note the chipset temps... before replacing the TIM they were at roughly 50C idle :yepp:

This may be only 1 hour, but it's a lot better than what I was getting before :D

http://i293.photobucket.com/albums/mm57/LennyRhys/4105p95.png

And I'm tempted to go a little higher, maybe if 1.35 vcore is enough for 4.15GHz - memory bandwidth is sweet at 1175, 5:6.

http://i293.photobucket.com/albums/mm57/LennyRhys/41561174PL8.png


EDIT: OK I'm going to settle for 4.1 with 483x8.5 because for the small increase to 4.15GHz I need 1.375vcore which is far too high for my taste. 4.2GHz might be a faster clock but I prefer high fsb/low multi. Thanks for the help everybody. :up:

[In-Fluence]

Are your CPU GTLs tuned? I think you'll find at higher fsb that this reduces the vcore needed for stability. Try 0,-45,0,-45 to start with and increase / decrease all values by 10, making note of the time it takes to freeze or give errors.

I haven't had this board long, but after more research and testing, I'm finally able to go higher than I did on my P5B thanks to tuning the CPU GTLs with only small increments in vcore. I'm currently testing 466x8 on my chip but with a 65nm Q6600 which needs +40,-5,+40,-5 (the values are inverted for 65nm). Results will be posted soon, promise :D

[Grnfinger]

the secret to this board is GTL's setting auto values is not really a good idea, one persons values will not always work for another its better to find your own.
I have found that anything above 470FSB requires a large increase in volts to keep it stable, the board really shines at 500+fsb

http://i237.photobucket.com/albums/ff267/Grnfinger/500x8_1200ram.jpg

[LennyRhys]

I will try to tune the GTLs soon because I really want to take the vcore down. All I have so far is 1hr of prime blend which isn't really much, but it's a better start than I had a few days ago lol. :D

[CryptiK]

My new 8600 seems to be pretty decent, I'm in two minds as to whether or not to give this thing run on the M2F.

Max boot and SS on air @ 1.552 vcore (idle)

5062 MHz

http://i720.photobucket.com/albums/ww206/CryptiKone/5062MHz1552vair.jpg

Max SPI 1M on air @ 1.552 vcore (idle)

5035 MHz

http://i720.photobucket.com/albums/ww206/CryptiKone/SPI1M5035MHz1552vair.jpg

[WaterFlex]

Hmmm, too much volts for 5Ghz E8600. My E8400 runs at 500x9 @ 1,34 volts. (LinPack stable)

[CryptiK]

It's 1.552v idle, 1.520v load. Not the best E8600 but it seems to be scaling decently. What a cpu will do at 4 - 4.5 GHz does not really indicate what it will do at higher speeds like 5GHz +. For example some cpu's that run stable at 4 GHz @ 1.15v can't even run SPI1M at 5035 MHz @ 1.55v ;) This chip is also very hot, the VID is 1.1125v.

Give yours a go, would be good to have another comparison as to what it needs to boot and run SPI 1M @ 5GHz +

[WhiteFireDragon]

cryptik, that's actually a really nice e8600. VID 1.11 is godly low. 5g on air is very respectable. what kind of tuning did you do for these clocks? BIOS template?

EDIT: i just realized this was not on the m2f, but DFI board. can you try on the m2f and post back on this thread? i'm curious if this crap board can hit those same exact clocks

[CryptiK]

Thanks yeah I hope it keeps scaling with more voltage and cold (when I get my gear). Took those SS's on my DFI X48 board, but VCore was 1.5750v (BIOS set) giving 1.552v (Real idle) and 1.520v (real load), Vtt = 1.30v, both CPU GTL Refs set to 0.667x, PLL = 1.51v.

[WhiteFireDragon]

what voltage can you get with it at 4500 prime stable? well good luck with your chip :up:

[CryptiK]

No idea, I got this cpu to just bench with so I haven't stability tested it.

[In-Fluence]

Go on give it a try in the M2F :)

Would be nice to see what kind of FSB you can reach too. I plan to do the same with an E8400 once I have the quad tuned. Managed an almost stable 550fsb with it on my P5B-dlx, and that was before I got the rig watercooled :)

[jason4207]

It's 1.552v idle, 1.520v load. Not the best E8600 but it seems to be scaling decently. What a cpu will do at 4 - 4.5 GHz does not really indicate what it will do at higher speeds like 5GHz +. For example some cpu's that run stable at 4 GHz @ 1.15v can't even run SPI1M at 5035 MHz @ 1.55v ;) This chip is also very hot, the VID is 1.1125v.

Give yours a go, would be good to have another comparison as to what it needs to boot and run SPI 1M @ 5GHz +

That is a low VID. Are you sure it's not the min-VID? Use RealTemp-->Settings to display the max-VID and min-VID.

[Oki]

Hi guys, i've finally been accepted in your forum so now I can thank you all for your generous contributions in my overclocking adventure with the MaxFormII, especially the simps guide to finetune all the settings and it also took me a better processor for getting higher clocks because my first processor was a E8400 C0 and the max stable was 3880Mhz even at 1.47V. And with my new E8500 E0 I'm running 4302Mhz 24/7 at 1.38V it will even go 4555Mhz at around 1.42V. But I'm very happy with the 4302Mhz. I'll try to post a screen of everest too.
http://img198.imageshack.us/img198/8135/superpi32m658235.th.jpg (http://img198.imageshack.us/i/superpi32m658235.jpg/)

Could you tell us what was the room temperature because it is pretty unreal that your idle temps are 25 !? Also what cooler is on that E8500 ? True ? ifx-14 ?

:welcome:

[Haes]

Hi Guys,

I appreciate this thread and not to mentioned simps' awesome guide (thx for that, dude - did 20 Mhz FSB more from 470 to 490 stable with your work).
I've a Q6600er G0 and want to buy a Q9650in a short time, but first I wanted to test the M2F. Normally I run this couple with 4x2Gbs 8500er Dominators from Corsair by 3,2 Ghz / 4*800 and 1.27V VC, 1.5V PLL, 1.1V NB, 1.1V VTT, 1.5V/1.1V SB with CPU/NB GTLs all +30mV rock stable - rest auto.

Bios 1307


This is the Setup for 490:

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 6
FSB Frequency : 490
CPU Clock Skew : Delay 100ps
North Bridge Clock Skew : Delay 300ps
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-980
Dram Clock Skew CH1 A1 : Delay 50ps
Dram Clock Skew CH1 A2 : Delay 50ps
Dram Clock Skew CH1 B1 : Delay 50ps
Dram Clock Skew CH1 B2 : Delay 50ps


DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : Auto
Row Refresh Cycle Time : Auto
Write Recovery Time : Auto
Read to Precharge Time : Auto

Read to Write Delay (S/D) : Auto
Write to Read Delay (S) : Auto
Write to Read Delay (D) : Auto
Read to Read Delay (S) : Auto
Read to Read Delay (D) : Auto
Write to Write Delay (S) : Auto
Write to Write Delay (D) : Auto

Write to PRE Delay : Auto
Read to PRE Delay : Auto
PRE to PRE Delay : Auto
ALL PRE to ACT Delay : Auto
ALL PRE to REF Delay : Auto

[like bios shows]

DRAM Static Read Control: Disabled
Dram Read Training : Disabled
MEM OC Charger : Enabled
Ai Clock Twister : Moderate
Transaction Booster : Manual

Common Performance Level [8]

Pull-In of CHA PH1 Disabled
Pull-In of CHA PH2 Disabled
Pull-In of CHA PH3 Disabled
Pull-In of CHA PH4 Disabled
Pull-In of CHA PH5 Disabled
Pull-In of CHB PH1 Disabled
Pull-In of CHB PH2 Disabled
Pull-In of CHB PH3 Disabled
Pull-In of CHB PH4 Disabled
Pull-In of CHB PH5 Disabled

PCIE Frequency : 100

CPU Voltage : 1.24
CPU PLL Voltage : 1.57
FSBT : 1.37
DRAM Voltage : 1.8
North Bridge Voltage : 1.4
South Bridge Voltage 1.5 : 1.5
South Bridge Voltage 1.1 : 1.1

CPU GTL Reference 0 : -25mV
CPU GTL Reference 1 : 20mV
CPU GTL Reference 2 : -25mV
CPU GTL Reference 3 : 20mV
North Bridge GTL Reference : +30mV
DDR2 Channel A REF Voltage : 12.5mV
DDR2 Channel B REF Voltage : 12.5mV
North Bridge DDR Reference : 25mV

Load Line Calabration : Enabled
CPU Sread Spectrum : Disabled
PCIE Spread Spectrum : Disabled


It's small, large, custom FFT and blend stable.

Now my case:

I hit with ease 500 FSB, but it crashes alway by about 10-15 minutes large fft or blend. If I bump 2 or 3 steps the voltages, same results. GTLs in Simps' green sectors didn't work, in fact -25/20 is my most stable setup.

Temps are all normal and good ... NB goes not over 55 °C in the stress tests.

Any hints ?

Thx

Keep up the good work ! :)

[In-Fluence]

Looks like you got yourself a nice o/c potential there, better than me anyway...

According to simp, NB & PLL voltages needed to be a lot higher than at 490fsb. try 1.63 pll, and at least 1..45v NB. Can't really say I'm talking from experience though, I only get 2 cores showing at 490fsb+ :(

[LennyRhys]

Bump the south bridge to 1.2; I noticed a HUGE increase in stability when making that jump. Your NB might need 1.45 for 500fsb, and the NB GTL may need to be retuned - at +50mv my overclock is fine, but one notch too high or too low and it doesn't hold... at +80mv all four cores fail within a few seconds. :yepp:

Another thing: try different PCI-E frequencies, I'm on 102 at the moment, but try 101, 103 etc.

[Grnfinger]

Hi Guys,

I appreciate this thread and not to mentioned simps' awesome guide (thx for that, dude - did 20 Mhz FSB more from 470 to 490 stable with your work).
I've a Q6600er G0 and want to buy a Q9650in a short time, but first I wanted to test the M2F. Normally I run this couple with 4x2Gbs 8500er Dominators from Corsair by 3,2 Ghz / 4*800 and 1.27V VC, 1.5V PLL, 1.1V NB, 1.1V VTT, 1.5V/1.1V SB with CPU/NB GTLs all +30mV rock stable - rest auto.

Bios 1307


This is the Setup for 490:

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 6
FSB Frequency : 490
CPU Clock Skew : Delay 100ps
North Bridge Clock Skew : Delay 300ps
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-980
Dram Clock Skew CH1 A1 : Delay 50ps
Dram Clock Skew CH1 A2 : Delay 50ps
Dram Clock Skew CH1 B1 : Delay 50ps
Dram Clock Skew CH1 B2 : Delay 50ps


DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : Auto
Row Refresh Cycle Time : Auto
Write Recovery Time : Auto
Read to Precharge Time : Auto

Read to Write Delay (S/D) : Auto
Write to Read Delay (S) : Auto
Write to Read Delay (D) : Auto
Read to Read Delay (S) : Auto
Read to Read Delay (D) : Auto
Write to Write Delay (S) : Auto
Write to Write Delay (D) : Auto

Write to PRE Delay : Auto
Read to PRE Delay : Auto
PRE to PRE Delay : Auto
ALL PRE to ACT Delay : Auto
ALL PRE to REF Delay : Auto

[like bios shows]

DRAM Static Read Control: Disabled
Dram Read Training : Disabled
MEM OC Charger : Enabled
Ai Clock Twister : Moderate
Transaction Booster : Manual

Common Performance Level [8]

Pull-In of CHA PH1 Disabled
Pull-In of CHA PH2 Disabled
Pull-In of CHA PH3 Disabled
Pull-In of CHA PH4 Disabled
Pull-In of CHA PH5 Disabled
Pull-In of CHB PH1 Disabled
Pull-In of CHB PH2 Disabled
Pull-In of CHB PH3 Disabled
Pull-In of CHB PH4 Disabled
Pull-In of CHB PH5 Disabled

PCIE Frequency : 100

CPU Voltage : 1.24
CPU PLL Voltage : 1.57
FSBT : 1.37
DRAM Voltage : 1.8
North Bridge Voltage : 1.4
South Bridge Voltage 1.5 : 1.5
South Bridge Voltage 1.1 : 1.1

CPU GTL Reference 0 : -25mV
CPU GTL Reference 1 : 20mV
CPU GTL Reference 2 : -25mV
CPU GTL Reference 3 : 20mV
North Bridge GTL Reference : +30mV
DDR2 Channel A REF Voltage : 12.5mV
DDR2 Channel B REF Voltage : 12.5mV
North Bridge DDR Reference : 25mV

Load Line Calabration : Enabled
CPU Sread Spectrum : Disabled
PCIE Spread Spectrum : Disabled


It's small, large, custom FFT and blend stable.

Now my case:

I hit with ease 500 FSB, but it crashes alway by about 10-15 minutes large fft or blend. If I bump 2 or 3 steps the voltages, same results. GTLs in Simps' green sectors didn't work, in fact -25/20 is my most stable setup.

Temps are all normal and good ... NB goes not over 55 °C in the stress tests.

Any hints ?

Thx

Keep up the good work ! :)

Why do you need dram skews when your underclocking that kit?
Even running 8GBs you shouldn't need skews

[In-Fluence]

I think if you follow simp's guide word for word, (i.e. set 266/333 NB strap, with 333 apparently being the most stable for high fsb) that's what haes ended up with. I took a close look at his settings though and thought I need to drop quite a few voltages as they were a lot higher than they really need to be.

Just out of interest Haes, could you find out what revision the chipset is? Mine's an A3 and although I would expect things to be very similar, my board has misbehaved from day 1 and I'm trying to ascertain if the chipset revision might be responsible...

[Stormad]

Could you tell us what was the room temperature because it is pretty unreal that your idle temps are 25 !? Also what cooler is on that E8500 ? True ? ifx-14 ?

:welcome:

Hey Oki, I think the roomtemp was about 22°C but in realtemp the temperature reads always to low, in coretemp the temp is 5°C higher so that would be more correct: 30°C, the cooler is a TRUE push/pull fans.

[Grnfinger]

I think if you follow simp's guide word for word, (i.e. set 266/333 NB strap, with 333 apparently being the most stable for high fsb) that's what haes ended up with. I took a close look at his settings though and thought I need to drop quite a few voltages as they were a lot higher than they really need to be.

Just out of interest Haes, could you find out what revision the chipset is? Mine's an A3 and although I would expect things to be very similar, my board has misbehaved from day 1 and I'm trying to ascertain if the chipset revision might be responsible...

What kind of problems are you having???

[In-Fluence]

I think its ok now, after switching back to 1307 and loading a profile that was tested 400*9 stable, but with ram on a lower, undesired divider (cant remember speed, maybe 1000MHz).
So I basically upped the speed to 1200 and PL8 and went from there. Setting the CPU/NB clock skews to normal/400 has pretty much solved the problems I was having, and after experimenting at 466x8 before, I got the same result, but kept hanging when a small fft test would reach 12K size (I did a custom test too, and it kept failing specifically on 12k). For some reason I can't get the board to improve over 466fsb.
At one point, and probably by pure chance I was able to post & even prime at 500 for an hour or so, but only 2 cores were detected by the board no matter what I tried. Odd behaviour imo, but I'm just going to call it a quirk for now. Anyway, I digress...

So far, I'm memtest stable, small fft stable and have blended for over an hour until prime quit with normal/400 skews. All other skews particularly the narrower ones resulted in instant failure by either not posting, or failing blend in under a minute. This is without any CPU GTLs tuned either which I hope will let me go a bit higher.

So from where I am now, and providing I get it stable overnight with a little more vmch or testing NB GTLs, If I then lower vCore to make it unstable, would you agree that this would be a good way to search for the right CPU GTLs?
If I then succeed in finding them, would upping vcore afterwards to get more FSB mean I would need to tune them again? I'm assuming not as it's more tied in with vtt (Which is at 1.3v).

I plan head up to 410x9 (the same as I used to run on the P5B) but stretch the memory more toward the 1230+ mark and tune the ram skews, and NB/DDR reference voltages.

After this, I'll call it quits until I spend some time using the bloody thing! Then I might go for a bit of a burn in 2 months or so after the summer cools off a bit :)

Thanks for offering your help Grnfinger, and others' input, it's always appreciated.:up:

*update
I upped the NB voltage and as I guessed, it gave me overnight stability. So I'm now tuning the CPU GTLs and so far, I've discovered that you were right grnfinger. the gap of 40mv as mentioned in simp's guide doesn't seem to give me the best results. The best result with the 40mv gap (50,10,50,10) gave me around 18 mins of stability, while the best with no gap (so far 10,10,10,10) gave around 45mins. A pattern does eventually emerge and I'm probably going to try some smaller gaps to see if I can home in on the longest lasting combination.

After that will be the NB GTL. System prime blended at DDR1200 PL6, strong on 1.37 NB volts for 30mins (which I stopped) but couldn't post at anything lower. To find a good unstable test point for this, would you suggest lowering the NB voltage or perhaps going for stronger twister and lowering the PL to 8?
If I up the FSB even by 1MHz, prime is more or less immediately unstable. Adjusting the ram skews only helped give a few seconds longer too, but wasn't sure if it was a good point to tune them so I returned them to auto.

[DaLiu]

Hi guys, my Rampage Formula died yesterday, some strange error after checking devices, USB overpower detected, shutting down in 15 seconds , didnt manage to find the fix so i need to send to RMA so i order a new Maximus 2 Formula, can you please recommend me some settings for Q9450, i would like to manage 475 x 8, on Rampage i manage only 450 x 8, require too much vNB for higher fsb. Probably i will try bios 1901 for my new Maximus 2 Formula and for the memory i will use OCZ DDR2 PC2-9200 Flex II 4GB, so please any advise or some settings will be helpfull for me as i dont understand this things with GTL for NB and CPU. Really appreciate your support!

[Kayso]

Hi guys, my Rampage Formula died yesterday, some strange error after checking devices, USB overpower detected, shutting down in 15 seconds , didnt manage to find the fix so i need to send to RMA so i order a new Maximus 2 Formula, can you please recommend me some settings for Q9450, i would like to manage 475 x 8, on Rampage i manage only 450 x 8, require too much vNB for higher fsb. Probably i will try bios 1901 for my new Maximus 2 Formula and for the memory i will use OCZ DDR2 PC2-9200 Flex II 4GB, so please any advise or some settings will be helpfull for me as i dont understand this things with GTL for NB and CPU. Really appreciate your support!

Here you go :)


Processor: Q9450
Ai Overclock Tuner: Manual
OC From CPU Level Up: Auto
Ratio CMOS Setting: 8
FSB Frequency: 475
CPU Clock Skew: Auto
NB Clock Skew: Auto
FSB Strap to North Bridge: 333
DRAM Frequency: DDR2-1140MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

1st Information:

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 15 DRAM Clocks
RAS# to RAS# Delay: 3
Row Refresh Recycle Time: 50 DRAM Clocks
Write Recovery Time: 6
Read to Precharge Time: 3

2nd Information:

Read to Write Delay (S/D): Auto
Write to Read Delay (S): Auto
Write to Read Delay (D): Auto
Read to Read Delay (S): Auto
Read to Read Delay (D): Auto
Write to Write Delay (S): Auto
Write to Write Delay (D): Auto


3rd Information:

Write to PRE Delay: Auto
Read to PRE Delay: Auto
PRE to PRE Delay: Auto
All PRE to ACT Delay: Auto
All PRE to REF Delay: Auto
DRAM Static Read Control: Enabled (setting it to Enabled gave me some more preformance in Everest)
DRAM Read Training: Enabled (and some more preformance too)
MEM. OC Charger: Disabled
Ai Clock Twister: Stronger
Ai Transaction Booster: Manual

Common Performance Level: 8
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
Pull-in of CHA PH3: Disabled
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled
Pull-in of CHB PH3: Disabled

PCIE Frequency: 101

CPU Voltage: You have to find your best one
CPU PLL Voltage: 1.55
FSB Termination Voltage: 1.325
DRAM Voltage: again you have to find your best
North Bridge Voltage: 1.45
South Bridge 1.5 Voltage: 1.55
South Bridge 1.1 Voltage: 1.15

CPU GTL Reference (0): Auto
CPU GTL Reference (1): Auto
CPU GTL Reference (2): Auto
CPU GTL Reference (3): Auto
NB GTL Reference: you have to find what will be your best here - start from +20 to +70 mv.
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 8
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled

Load-Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

[DaLiu]

Guys i just receive my M2F but i have one question, my temps are quite high for NB and SB and i already change the :banana::banana::banana::banana:ty paste with MX2 under NB, SB and mosfets, the truth is quite hot here, around 30 C in the apartment, is so hot but do you think are normal this temps, usually for my last Rampage Formula i used Thermalright HR5 SLI for NB and SB and for 450 fsb with 1,52 v in NB the temps in full load was around 58 for NB and 52 for SB.

http://www.myrmidoni.com/_stuff_/temps.JPG

Edit later: after 5 loops of IBT the temps for NB - 60 C and for SB - 56 C, only 4-5 from idle to load.

[orbiter]

Guys i just receive my M2F but i have one question, my temps are quite high for NB and SB and i already change the :banana::banana::banana::banana:ty paste with MX2 under NB, SB and mosfets, the truth is quite hot here, around 30 C in the apartment, is so hot but do you think are normal this temps, usually for my last Rampage Formula i used Thermalright HR5 SLI for NB and SB and for 450 fsb with 1,52 v in NB the temps in full load was around 58 for NB and 52 for SB.

http://www.myrmidoni.com/_stuff_/temps.JPG

With ambient temps that high I would say your NB/SB temps are fine. Just to be on the safe side I always set NB/SB overheat protection temps down to 70*C in BIOS in the hope that this would give early warning of temperature troubles.

[In-Fluence]

I agree, your temps are fine, here's a snapshot of my idle temps in a room at 25°C, bearing in mind I also have a waterblock on the CPU and above the NB:
http://i194.photobucket.com/albums/z182/in-fluence/Temps25Cambient.jpg

With regards to my rig, I now have the best CPU GTLs set to +10 for all 4, and got the memory to memtest overnight (26 passes) at 5-5-4-15 timings on 1200MHz PL6 strong, with the CPU at 400x9, but now blend test is failing a lot and have adjusted NB voltages to 1.33, 1.37 and 1.4 whilst testing NB GTLs at each voltage, but nothing seems to give any hints for stability, all combinations give rounding errors around the same time (give or take a minute)..The most successful I noted was with NB at 1.33 and NBGTL at +60 which gave me 5 mins of stability and prime quit, 1.37 with +70 gave me rounding errors on 2 cores but kept priming the other 2 for a while longer before I quit it.

Am I going the right way about this?

[DaLiu]

Thank for info, i made a mode on my mobo, i put a Thermalright HR-5 SLI on NB, i will put the pictures soon, right now the temps for NB are fine, 55 C with 5 loops of IBT, in idle 51 C but look like the biggest problem for me is that the videocard is heating the SB which is in idle 54 C and for sure is heating the NB too. Ill start with overclocking as soon is not gonna be so hot.....ahh btw i just broke the lcd poster, basically i just wanted to take it out and i brake the cable, good quality, usually on my Rampage Formula i took it out couple of time and no problems, on M2F first time and is already broken, too bad it could help when i start to overclock.

[Grnfinger]

download everest to monitor temps and volts with, the lcd poster is rather limited in its functions

[DaLiu]

For the moment i manage to reach 450 x 8, for 475 x 8 the cpu need looks like quite a lot of juice and with my cooling i cant affort to give it, for 3600 Mhz is going with 1,3 V but for 3800 Mhz 1,33 V is not enought and i reach in IBT around 88 C with 30 C in the room, ill buy soon a proper cooler and after i can give more juice to my cpu, for the moment here are some pictures from today.

http://www.myrmidoni.com/_stuff_/17.07.2009/1.jpg
http://www.myrmidoni.com/_stuff_/17.07.2009/2.jpg
http://www.myrmidoni.com/_stuff_/17.07.2009/3.jpg
http://www.myrmidoni.com/_stuff_/17.07.2009/4.jpg
http://www.myrmidoni.com/_stuff_/17.07.2009/5.jpg
http://www.myrmidoni.com/_stuff_/17.07.2009/6.jpg

[jason4207]

Am I going the right way about this?

Yes. It's going to be hard to get stable there, though. I would just use large-fft for now since you're pushing the FSB/RAM system so hard. It'll fail faster (or at the same time) if there is a problem in that area.

If you're only 5mins stable at the max it's going to be hard to find something that stays 24/7 at those speeds/timings. You might have to put PL to 7.

I fought the same battle for weeks, and finally settled for a little less.

You might want to also try for 450x8 4:5 1125 PL7-8 Strong 5-5-4-x

[Haes]

Sorry Guys,

my internet is off since Tuesday cause I switched the provider and there's a small problem that will be fix tomorrow.

Thanks for the help.

@ In-fluence: It's A2 Revision and the first few times I had a lot of problems too - compared to my old Asus MB (p5k). The secret is a well tuned skew table, like simps and others mentioned it here already and the m2f will pay it with stability by less voltages.

To the RAMskewtopic:

Indeed I don't have to tune the skew when I underclock the RAM - I just wanted to be sure. The 333 Strap is the best avaible at the m2f -bring the most stable setup from my experience with this board.

Set the PLL to 1,6V is for my taste a bit to high for a 24/7 Setup - even if it's a 65nm Quad. But for testing some of your hints, especially the SB Voltages and PCI-E Frequenzies -I'll give it a try. Like I said I can boot and work quite a time by 500 FSB but stressing it , will end in the BS in 15 minutes.

But like i wrote, few days ago I couldn't even get 470+ FSB stable and now i'm at 490 atleast and can boot and work a bit in 500. So or so, it's an improvement thx to you.

Best Regards

Haes

[In-Fluence]

I've been away for the weekend so testing was postponed until last night. I went back to 450x8 3:5-1200 for the sake of running a higher memory bandwidth (and for some reason the CPU idles a few degrees cooler with same voltage). I used the same settings as I did with 400x9 but I needed a bump in NB volts to 1.41 to get it stable'ish' (left it testing, but at work atm), but I'm going to take a step back and re-check the CPU GTLs as well as I might expect different values could be needed for higher FSB and the whole process could be more sensitive when testing too. A lot of my last GTL tests gave rounding error results quite close to each other and nothing really 'stood out' for me.

I want to try for 1200-5-5-4-15 on the 3:5 divider but I guess that could prove problematic. Failing that I will try the 4:5 divider and go for tighter timings. I'm glad to know the memory is capable of performing quite well anyway. It seems the northbridge is the trickiest thing to try & tame, and mine seems to like a bit more voltage than most going by settings I've seen on here, but gets worse when I start to go past ~1.45nbv. It's definitely taking a while but I am getting there and definitely have a better feel for the board now.

*update - went down to 4:5 divider as you suggested Jason, and seems a lot easier to get stable now whilst still giving the RAM enough headroom to try and head beyond my previous max CPU overclock. Currently benching at 457x8 @ ~1144 5-5-4-15. Would be really nice to eventually up the FSB to the 466 region and push the ram a teeny bit more. *end of update

I've also got my hands on an E8500 capable of 550+fsb (p5b's limit), so I'm eager to give this a whirl as my last attempt was on air too. I'm almost assuming getting this chip stable at higher speeds will be a breeze in comparison, just so long as I go steady on the volts. I guess the only snag could be the 1:1 divider which I have read doesn't work too well on this board. I should probably persevere with the quad until I get something good & rock solid actually, but I'm sure you know how it is when you get the o/c bug! :D.

[DaLiu]

Guys can you please tell me what kind of temps do you have for NB and SB on air please, if is posible on idle on load, thanks.

[In-Fluence]

Looking at your NB temp I would say that is a little too hot for comfort, but that depends on what voltage you are using for it. 65°C would probably be ok for ~1.55v, but if the temps you posted earier are for something like 1.3v then you will probably want to remove the heatpipe assembly and all the paste, clean, then re-paste the SB & NB and make sure you screw the NB down first (maybe with a couple of extra washers to add pressure), then tighten up the surrounding screws.

Are you sure you have that thermalright heatsink on flat against the mounting? It also seems like you've got a lot of paste on that too, it shouldn't really be coming out of the edges like that, you might find an improvement if you re-apply a smaller amount otherwise it insulates unwanted heat.

[Haes]

Not more than 56°C Prime95 stressing, stock TIM and no matter how much I push the NB against FSB 500 or 1,45V. Normally it idles by 45°C-48°C depending on the roomtemps and load by 50°C-53°C.

I'll change the TIM cause many users are shocked by the mess of loaded TIM on the NB which asus' manufacturing provides - meaning they put too much, and after the change they recognize significant tempdrops by about 5°C-8°C.

[CryptiK]

With temps of 65*C on your NB you really should redo the paste with something like ceramique or MX2. If you want some reference pics for the heat pipe removal and paste removal/application I documented it here:

Heatpipe Removal / New Thermal Paste Application for M2F (http://www.xtremesystems.org/forums/showpost.php?p=3374541&postcount=1511)

[DaLiu]

I already changed the paste under NB and SB and believe me that i desambled the mobo 2 times in 1 day just to try to fix the SB and NB how better i can, the problem i think is from the video card which is heating a lot the SB, even in idle the SB is 52 C and NB is 57 C, in load under IBT or after couple of hours of playin i reach 67 C for NB with 1,38 V and 62 C for SB. Soon im gonna change the cpu cooler and i need to take the mobo out of the case and im going to try to fix the radiator again on NB. I searched over the internet and there are many cases like me but with the NB on default voltage, so im thinking that the NB heatsink is not having enought pressure over the NB chip, i really dont know what could be the problem. In the room right now are around 28 C, probably inside the pc are 10 C more, the temp for the MB is 39-40 C. With opencase for 5 min the SB temp is going on 50 C in idle and NB on 56 C.


http://www.myrmidoni.com/_stuff_/temps.JPG http://www.myrmidoni.com/_stuff_/tempsload.JPG

First picture the temps in idle with the case closed and second picture is with the case opened and you can see the temps just before 5th loop of IBT to be finished, what do you think?

[Grnfinger]

57c is too hot for idle imo, my NB wont see 45c under full load and idles at a cool 35-37c and I run my NB at 1.45volts
you should try a spot cooler to help keep the NB under control

[In-Fluence]

My results so far:
http://i194.photobucket.com/albums/z182/in-fluence/365GHz-1144stable.jpg

And bandwidth scores (which I intend to work on):
http://i194.photobucket.com/albums/z182/in-fluence/365GHz-1144stable-everest.jpg

Definitely much easier for me to get stable at 450+ than at 400+ fsb. 460 is next :)

[WaterFlex]

In-Fluence
Asus Maximus II Formula Rev.A3???
What`s the revision?

[In-Fluence]

The board is a 1.02G, and the P45 chip itself has the stepping SLB8C. Not in a position to compare differences just yet but I hope it's likely to bring better results like they apparently have on the P5Q and Biostar i45 boards. So far my experiences haven't reflected this, but maybe after some more tuning things might improve.

[DaLiu]

Guys i have a question, i tried to find all over the internet any detail about the NB and SB temps and how they can be fixed, i think the biggest problem is the "legs" (dont know how to call them) from the heatsink (SB-NB-heatpipes), they are a little bit long and even if i use washers under the mobo basically i cant make the heatsink to put pressure on the NB only if i cut a little from the legs, i hope you understand me, and im sure this is the biggest problem in my case because i already desamble the heatsink from them mobo 2 times just to be sure that i put the MX2 correctly on NB and SB. I tried to find a better picture with the heatsink of M2F but i didnt find any from different view, basically there are the legs with some hard rubber over them, i think if i manage to take off that rubber i could put a better presure over the nb, sb and mosfets (even this are going on 60C in full load). What do you think, right now im waiting for Prolimatech Megahalems which is coming only in 10 days and when i have the cooler im going to try to fix the heatsink as i said.

http://www.overclock3d.net/gfx/articles/2008/07/08142942135l.jpg

[In-Fluence]

Looking at that, I think we can safely assume your NB is definitely not making good enough contact. Another thing to note is the PWM heatsink on the left - only half of those are making contact, unless the photo / lighting is making it look that way. If the 'legs' are easy to remove, I would suggest trying to remove the ones either side of the NB area and putting a thinner washer in its place. I would leave all the others as temps you have reported everywhere else seem perfectly normal.

I'm a tad disappointed with the consistency of asus build quality - very poor from the number of incidents related to this heatsink assembly so far. I think the whole thing is over elaborate and am tempted to remove the whole thing myself after I've got my ideal o/c :(

[DaLiu]

For the moment ill keep like this for 1-2 weeks till ill receive my new cooler and after that i will desamble the hole system and i will try to fix this :banana::banana::banana::banana:ty heatsink properly. Right now with the ambiental temp quite high, is very hot on my location the biggest temps i saw on NB and SB in full load was 67 C for NB, 60 C SB and 62 C for mosfets. Worst than this could not be.

[In-Fluence]

Mosfets & SB are fine in comparison to mine. The SB does run hot, but it can take it, and probably a lot more. I think you'll find it's quite a different chip to the 45nm NB, hence the higher voltage & temps. The mosfets are also normal, trust me, I've put some aftermarket pure copper sinks on them and the temp only dropped by 1-2°C, they will run hot on most boards from my experience.

I will have to test the difference between having the Asus heatsink system on there and having 'all seperate' sinks and see how they compare. Although I didn't notice anything major the last time (I think the SB ran hotter), I think it would be interesting to see if any heat from one area has a significant impact on others.

[DaLiu]

I think it would be interesting to see if any heat from one area has a significant impact on others.

I think the SB is heated more than usual in my case because im using accelero xtreme 8800 on my 8800GTX and a lot of hot air is going directly to SB and to NB for sure, ill change the video soon and probably ill see some improvement in the temps area, the only problem is if is posible to take off that hard rubber from the legs of the headsink and i hope to not damage anything.

[Grnfinger]

Guys i have a question, i tried to find all over the internet any detail about the NB and SB temps and how they can be fixed, i think the biggest problem is the "legs" (dont know how to call them) from the heatsink (SB-NB-heatpipes), they are a little bit long and even if i use washers under the mobo basically i cant make the heatsink to put pressure on the NB only if i cut a little from the legs, i hope you understand me, and im sure this is the biggest problem in my case because i already desamble the heatsink from them mobo 2 times just to be sure that i put the MX2 correctly on NB and SB. I tried to find a better picture with the heatsink of M2F but i didnt find any from different view, basically there are the legs with some hard rubber over them, i think if i manage to take off that rubber i could put a better presure over the nb, sb and mosfets (even this are going on 60C in full load). What do you think, right now im waiting for Prolimatech Megahalems which is coming only in 10 days and when i have the cooler im going to try to fix the heatsink as i said.


Have you tried adding extra washers to the back side of the board to increase the pressure?

[DaLiu]

Have you tried adding extra washers to the back side of the board to increase the pressure?

Not yet, but as i said i think the biggest problem in putting the pressure over the NB is that the legs are quite long and that rubber is to hard, is not banding, dont know if you understand me, my english sucks. Maybe im wrong, ill check it when i get the new cooler to do everything one time, right now i dont have the washers, they will come with the cooler, in the same order.

[jason4207]

Mosfets & SB are fine in comparison to mine. The SB does run hot, but it can take it, and probably a lot more. I think you'll find it's quite a different chip to the 45nm NB, hence the higher voltage & temps. The mosfets are also normal, trust me, I've put some aftermarket pure copper sinks on them and the temp only dropped by 1-2°C, they will run hot on most boards from my experience.

I will have to test the difference between having the Asus heatsink system on there and having 'all seperate' sinks and see how they compare. Although I didn't notice anything major the last time (I think the SB ran hotter), I think it would be interesting to see if any heat from one area has a significant impact on others.

I'm pretty sure the NB is 65nm.

P45 - 65nm

P35/X38/X48 - 90nm

[In-Fluence]

I stand corrected :)

[Haes]

Back again,

@ Lenny-Rhys:

So, I just use your advice bumping the SB voltage up to 1.2V and while I still can't get the 500+ FSB primestable at 1.45V NB, 1.4V VTT and 1.6V PLL, I could drop some voltages at 490 (here the first picture of it - watching l.a. confidential and writing this post).

It feels so stable, everything is flawless, no lagging , nothing. I'll do some adjustements and testing it again by less voltages on PLL, NB and VTT.

And it seems too, that I underestimated my 1,3125 VID G0. Never testing it at about 3 Ghz, but 1.19V VCore is nice (ok - it's blend test) :) .

http://img156.imageshack.us/img156/3956/testblend4901.th.png (http://img156.imageshack.us/i/testblend4901.png/)
after 1 hour

http://img229.imageshack.us/img229/8444/testblend49011.th.png (http://img229.imageshack.us/i/testblend49011.png/)
after 2,5 hours



Still nothing , passed 768k, no 12k...

By the way my 2nd pair DDR2 8500 Dominators from Corsair just said goodbye. One of them do some heavy mistakes while testing it with Memtest 3.86, even I never oced them over 1066 Mhz.
I've to replace them by Corsair (2 months old) and so this test is only by 4 GB RAM.

Plus I'll want to see how high I can get the FSB by 1.45V NB, 1.4V VTT and 1.6V PLL to boot Win7 and validate it by CPU-Z.

And the RAM Timings and so on aren't optimized/done yet.


So thanks for now :up: .

P.S. 12k passed -> 14k ...


*** update ***

did some test by reducing VTT/NB/PLL ... CPU GTL 20/-25/20/-25 NB 30

FSB 490:

1.25/1.25/1.51 Winboot Freez , no BS
1.29/1.29/1.51 Sysfreez by LargeFFT 15min, no BS
1.29/1.29/1.53 Sysfreez by LargeFFT 25min, no BS
1.32/1.32/1.53 Sysfreez by LargeFFT 25min, no BS
1.34/1.34/1.53 Sysfreez by LargeFFT 25min, no BS
1.36/1.36/1.55 all stable LargeFFT,Small FFT,Blend,Custom - did 3 hours without errors - just a first test

Strange, that it's such an increase on voltages, because even by 1.29V thre is no lagging etc. , perhaps have to readjust the GTLs and skews though in the last tests there was no improvement to see against de 20/-25/30 setuo.

Did just a spontaneous validate test

FSB 520 validated by 1.4 VTT, 1.45 NB, 1.55PLL

http://valid.canardpc.com/show_oc.php?id=611876.

But as you can see there are only 2 cores active. So I must be at my sys max. Did a short blend test on 520 for 1 hour and it worked without errors though cpu was at 3.12 Ghz and I forgot to set the Vcore right :D .

[humeyboy]

I'm pretty sure the P45 NB is 45nm going by lots of reading in the past and also on guides for the NB GTL's Tweaking.

The SB can be hotter for many reasons but also as its encased in black resin (like the old NF2) unlike the NB which has a bare die.

[DaLiu]

Guys can you please check something for me because i have a feeling that my temp sensors for NB and SB are reporting with around 10 C more. Basically after my PC was shutdown for around 8 hours i start the pc and right away i checked the temps in the BIOS, MB - 27, CPU - 28, NB - 38 and SB - 36 just after 10 seconds i pushed the button to start the computer, do you think is posible to rise the temp for NB and SB so fast or is a bad reporting? Because if is a bad reporting basically my temps are fine. Any help will be much appreciated, thanks guys!

Could you please check you temps for NB and SB in bios right away after you start you computer (being shutdown for couple of hours) just to check what are your tems just after you start the computer comparing with mine. Thanks!

[LennyRhys]

I'm getting roughly 50C load on NB and SB with the stock cooler, ambient temp is about 20C.

@ Haes - glad you had some success with 1.2 SB! :up:

[DaLiu]

I'm getting roughly 50C load on NB and SB with the stock cooler, ambient temp is about 20C.

@ Haes - glad you had some success with 1.2 SB! :up:

Can you check it please directly in the bios right away after you start the computer (not using it for couple of hours) just to see what is the starting temps for NB and SB in your case because in my case they are starting looks like from 38 for NB and 36 for SB and usually it supose to be around 27-28 C, im thinking that my mobo is not reporting corectly the SB and NB temps, for MB, CPU and PWM looks like the temps when i start the computer after not using for couple of hours are around 28 C and for the NB and SB are with 10 C more.

[jason4207]

I'm pretty sure the P45 NB is 45nm going by lots of reading in the past and also on guides for the NB GTL's Tweaking.

The SB can be hotter for many reasons but also as its encased in black resin (like the old NF2) unlike the NB which has a bare die.

Nope. Just google '65nm P45' and it's all over the place. Here's a good example:

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3325

There is no 45nm NB chipset that I'm aware of. The reason the P45 is different is not b/c it's 45nm, but b/c it's 65nm, and all other chipsets are 90nm.


Guys can you please check something for me because i have a feeling that my temp sensors for NB and SB are reporting with around 10 C more. Basically after my PC was shutdown for around 8 hours i start the pc and right away i checked the temps in the BIOS, MB - 27, CPU - 28, NB - 38 and SB - 36 just after 10 seconds i pushed the button to start the computer, do you think is posible to rise the temp for NB and SB so fast or is a bad reporting? Because if is a bad reporting basically my temps are fine. Any help will be much appreciated, thanks guys!

Could you please check you temps for NB and SB in bios right away after you start you computer (being shutdown for couple of hours) just to check what are your tems just after you start the computer comparing with mine. Thanks!

Chips can heat up in an instant and cool down in an instant. I seriously doubt your sensors are messed up. It's more likely due to your high ambient temps, and/or bad HS contact. You may also be using too much TIM by the looks of your pic above.

[Grnfinger]

I would agree
after a closer look at the nasty glob ooozing from the NB I would say there is way to much paste, its doing more harm than good

[DaLiu]

I would agree
after a closer look at the nasty glob ooozing from the NB I would say there is way to much paste, its doing more harm than good

Is not much, is just a thin layer but because the size of the thermalright hr-5 is smaller than the size of stock heatsink you can see some paste on the edge, anyway i will do the job again next week and i will clean everything and do it again with a very thin layer.

Anyway for my curiosity please check the temps for NB and SB in the bios when you turn on your computer.

[LennyRhys]

My "cold boot" chipset temps are in the low 30s. When I got to windows my NB temp was 35C and climbing.

It sits at 46C when all my case fans are on; with the top fan turned off (I accidentally snapped a cable :D) my NB temp climbs to over 50C, which shows just how much a big exhaust fan helps case airflow. :up:

[Grnfinger]

mine are in the 22-24c range when I power up my rig sits in a 17c room tho so I'm not really one to compare with
I'm also cooling NB and SB with water

[Blue Storm]

hello guys...tell me please, what are your best FSB frequencies with C2D E8x00...I want to buy this board, but I am sceptic, because I didn't find much high FSB results...very thanks Blue Storm

[eSp!s0]

I am stable at 500FSB with my E8400 (still trying to get 533 stable). But with the X38 MF I didn`t get the CPU booting above 510FSB, with the M2F I already booted with 550FSB.

[humeyboy]

On a Dual 500FSB should be easy, I had it on a E8600 with 8GB.

I don't know why I did not try more but probably as I had my heart set on jumping to 600FSB but would need 1200MHZ Memory (I can boot mines to 1175 with all 8GB/4 Modules with no fancy tweaks or voltages but still the same with only 2GB/1 Modules fitted, so assume its lack of Bios support).

Other peeps here have the same Rev 1.1 Memory and have had it over 1200mhz in other Mobos but hit same 1175mhz wall on this Mobo.

I did this with very low voltages, NB was 1.2v (default is 1.1).

I now only last night got 500FSB with the Quad (not stable yet).

Again lowish voltages even leave on Auto (apart from PLL which is too high on Auto), its all about getting the CPU and NB GTL Voltages correct, the killer is positive values are in 10's but minuses are in 5's so give better finer adjustment.

Before this I could not even Save and Exit the Bios at 500FSB, simply put my monitor into standby.

I have a sore head as of sitting up tp 9am messing with different Bios's and settings but there is not 1 thing I can say for sure apart from above because it would work with settings then not work with them next time, then work again.

Very strange and temperamental.

The only thing I have not done due to laziness and fact I have a cool running NB/SB as a later batch (still A2) is redo the TIM.

I would ask anyone here if they redid the TIM not due to heat (had normal temps) and found it actually held stabilise settings they were not stable before.

[Blue Storm]

I received this board today, but I have problem with sound card...it is connected on board, but Windows can't find it...It shows, I have no sound card...my system is Windows Vista Ultimate 32bit...do you know, what is the problem?

[humeyboy]

I do not use the includes Soundcard so sorry no clue and ain't seen/heard about any real issues.

[DaLiu]

I received this board today, but I have problem with sound card...it is connected on board, but Windows can't find it...It shows, I have no sound card...my system is Windows Vista Ultimate 32bit...do you know, what is the problem?

Check in BIOS if on board audio is enabled.

[Kayso]

I received this board today, but I have problem with sound card...it is connected on board, but Windows can't find it...It shows, I have no sound card...my system is Windows Vista Ultimate 32bit...do you know, what is the problem?

Have you connected it to the right port ? Because if not, it doesn't work at all :)
Look at the attached image.
http://img268.imageshack.us/img268/3558/maximuss.th.jpg (http://img268.imageshack.us/i/maximuss.jpg/)

[koksy]

I received this board today, but I have problem with sound card...it is connected on board, but Windows can't find it...It shows, I have no sound card...my system is Windows Vista Ultimate 32bit...do you know, what is the problem?
Did you remember to install the drivers?

[Blue Storm]

Check in BIOS if on board audio is enabled.

yes...in BIOS it's enabled, but it behaves as if it were not involved


Have you connected it to the right port ? Because if not, it doesn't work at all
Look at the attached image.

yes, I have put it there


Did you remember to install the drivers?

drivers can't be installed, because system can't find sound card

[humeyboy]

Have you connected it to the right port ? Because if not, it doesn't work at all :)
Look at the attached image.
http://img268.imageshack.us/img268/3558/maximuss.th.jpg (http://img268.imageshack.us/i/maximuss.jpg/)

It works in any PCI-E Slot BTW. ;)

[koksy]

drivers can't be installed, because system can't find sound card
Usually it's the other way around. You have to install the drivers before the system can see it.

[WhiteFireDragon]

It works in any PCI-E Slot BTW. ;)

no, the sound card that came with the board ONLY works in the first black slot. i've tried it in the other pci-e slots and it doesn't work.

[eSp!s0]

Usually it's the other way around. You have to install the drivers before the system can see it.

No, usually, the OS recognizes an audio device(uninstalled) and then you can install the drivers.

[kio]

Hi

Try to do my best with this Mobo

Limited by bad Ram
Try to do more next time

http://valid.canardpc.com/show_oc.php?id=633045

http://img23.imageshack.us/img23/1281/maxscreendice.jpg (http://img23.imageshack.us/i/maxscreendice.jpg/)http://img23.imageshack.us/img23/maxscreendice.jpg/1/w800.png (http://g.imageshack.us/img23/maxscreendice.jpg/1/)
http://img266.imageshack.us/img266/9898/p7090003.jpg (http://img266.imageshack.us/i/p7090003.jpg/)http://img266.imageshack.us/img266/p7090003.jpg/1/w1600.png (http://g.imageshack.us/img266/p7090003.jpg/1/)

[seban]

Hi guys,

Long time no see in this thread. I came by to say hello and also to answer a question asked long time ago here by someone - unfortunately i cannot remember who was asking ;).

Anyway. My q9550 running 4ghz and 24h prime blend stable / 20 passes linx / memtest 12h+ stable was resetting itself during idle - preferable when I was not in the room ;) (little bastard ;) - someone here had same problem - full stable according too all tests but resetting during idle.

In my case it was faulty OCZ 800w EliteXtreme PSU. It died about month ago (it was 8month old as whole rig). Because it was second from two products altogether from OCZ which was faulty which gives 100% of faulty items form OCZ for me - I decided never again to buy anything from them (ofc there was no problem with warranty in any case - and the company was very nice and helpful in both cases).

Coming back to subject I bought Corsair HX 850w and besides much less clutter in pc case due to modular design - guess what - during last month absolutely 0 restarts. So happy ;) and I feel really satisfied that I did not lower my OC when restart was ocuring.

I did not lower it because what is the point of changing anything if anyway all tools to test stability says that sys is stable - so what could I achieve by changing OC settings.

In the end it was PSU. Beware of hmm high-end PSUs ... such a shame ;)

Gl with OCing ppl

PS> Are there any dedicated drivers for our Mobo for WIN 7 no matter if they come directly from Intel or anyone else? I cannot find any? Anyone had more luck?

Thanks

[humeyboy]

I'm sick of this Mobo, will post full details when I am in the mood and not lazy.

Would seem all Asus Rog's are buggy as I have had a few Rog models, the lower models get better support.

[In-Fluence]

Spent some time today on a borrowed E8500 to see if the board was limiting me in any way. As you can see below, it isn't!

http://i194.photobucket.com/albums/z182/in-fluence/43GHz-1228PIBandwidth.jpg

I did have a good stab at trying to break 10 seconds 1M PI, but it couldn't quite do as well as my P5B which beat it by around .4 seconds. I think the CPU is limited to around 4.3GHz before silly voltages are required to push it further. I didn't tune any GTLs or skews either which could have helped more but I didn't want to waste lots of time on this. It was just nice to see 600+FSB.. I could even memtest at 630FSB but booting was gonna take a lot more tweaking.

So I've decided to go the same route as Seban and get me one of those lovely HX850's as soon as I have the money. This is now the only reason I can think of as to why I can't get the same point of stability as I did with my P5B - failing that, I think I might just stick with 3.6 and some nice low voltages.

p.s. great work woth your E8500 Kio, I bet your FSB could fly pretty high with that too!

[humeyboy]

Dual are easy on this Mobo its not so easy with Quads.

[In-Fluence]

You don't say? :) I do agree though - my quad seems 10x harder to tune and it's taken much much longer to get stable at any comparable overclock than on my P5B. That also had nothing like GTLs, Skews and PL settings which is what makes me think PSU unless there was a lot more going on 'under the hood' . I'm well in need of a new upgrade anyway (probably about 2 years overdue) and Corsair's HX850 seems like the perfect candidate.

[humeyboy]

A older 65nm Quad will be even harder but your lucky you have a newer A3.

I seen a review with 3GHZ & 500FSB on a Q6600 (No GTL Tweaks), makes me wonder if them 1st Mobos for review were golden chips same as that MSI 780 we seen at launch doing just under 700FSB with a Dual core and peeps were lucky to get 500FSB.

Can you post with your Memory above 1175MHZ (I mean even get out of Bios Screen) ?

That's some nice hardware you have but that old 65nm Quad CPU is the weak link (never did like them), would love to see a 45nm Quad on a A3.

[In-Fluence]

Yeh, my Q6600 is an '08 batch (L846A659), it seems the FSB is limited to around 475, around 476-499 is a 'dead zone' which wont post at all, and 500-515 only gives 2 cores. I also need to do normal/400 CPU/NB skew settings to get stable above 450, or post above 460. Setting the skews higher /further apart does let me get to around 475 but it's in no way stable 465 is the FSB I'm working on now, then I can hopefully jump up to 8x multi.

CPU GTLs are the settings I'm having the most trouble with (and would ask for any further advice if you have it). I've been through several different written pages worth of tests at different FSBs, VTTs, dividers and none of the values stand out that much, at least not like the skews did. From what I've read on this thread, GTLs have made a big difference for a lot of users, but that doesn't seem like the case for me. Each time I've tested them, then re-tested them at the same settings, they gave different periods of stability varying between 3 & 15 minutes! Rounding errors are still hard for me to provoke, but do occasionally pop up too (on high FFTs). This is where I'm most frustrated right now because there are soooo many combinations for the GTLs, and I get to a point where I think 'that's it' then it all goes out of the window after another failure or when I decide to try a different VTT. My last resort is to try some weirder settings like +40,+40,auto,auto. Perhaps this might do something...

Memory is happy to run at ~1233MHz on almost any divider (haven't tried them all) and the 1:1 ratio gave really nice results which could probably have been tightened a little more. The board behaved really nicely with the dually in there 'as I would expect it to', but never thought a 65nm Quad would play this much havoc with general stability. Memory runs a lot better on this than in the P5B which was the one factor that made me keep hold of it, although raw power is really what I'm after.

A couple more questions too if anyone could help out a little here:

1. To determine different instabilities in the form of rounding errors, hangs and reboots - would anyone say the severity of these are as listed? (rounding errors being most stable, and reboots being the least)

2. I've read about people pumping VTT higher than their Vcore in order to get stable on 65nm quads (as high as 1.6v!) I've had a little more success with this too but am VERY dubious and havent gone over 1.48v so far. I only needed 1.2 on the P5B (but running 410x9), do you think I should aim for above 1.3v or try to stay lower? I would like to stick around 465x8 for the bandwidth.

[humeyboy]

Try NB GTL's, made major difference from me but same as you above 475MHZ it works 1 reboot and will not next reboot.

Even had 500FSB benching but not for long as it freezes, I blame Bios and lack of Memory support as Corsair ain't even on the list.

Skews did nothing for me, but I cannot boot with a negative NB GTL but can at Auto or Positive.

Currently I am @ +40 NB for 475FSB and CPU GTL's are +20/+40/+20/+40, Tried to leave the 20's at Auto and also tried all Negatives but not sure what's best as of above info. because its simply random each reboot.

It makes no difference if I use 1 Module or all 4 Modules or even my older Corsair Modules, still hard to get over 475FSB and 1175MHZ seems limit though my Ram can do more.

My advise would be to sell of some of the non needed kit and buy a Q9650 and see how far it can go on an A3 after all that's what the Gigabytes are.

[seban]

post your template. But it is true that this MB starts to be really tricky over 470mhz.

[humeyboy]

Back at Auto/+40/Auto/+40 for now which means the Cores are not equal in the Priming and the longer they run the further apart they get, +20/+40/+20/+40 seems to fix this.

Currently (ongoing) ran a 13 hour Blend and now just about to complete a 2 hour Small and about to start a 2 hour Large, all stable with fans all at 50%.

Still a waste of time as next time I reboot it can all change.

I'm beginning to think my LLC is crap, on the NF780I it was solid but on this Mobo 1.2VCORE drops to 1.168v with it enabled but this stable run its at 1.192v (I have lowered Multi so that's not so important as I am testing FSB not overall MHZ).

NB and VTT Voltages also more solid this time around.

[In-Fluence]

I noticed how each cores testing times seemed to gradually move further apart at different settings, though I inititally thought it was down to background processes taking little bits here & there from each core - interesting. There is also talk about seeing CPU usage dip in task manager if GTLs aren't correctly tuned when on low speed update setting, but I haven't noticed it on that, (or with normal/high speed update), and speedfan doesn't do it either (even though that seems more responsive than task manager).

So far all on auto or +10 gave the most stability when testing at 466x7 which resulted in around 20 mins stability last night, but I woke up this morning and tested the same config and only got 3 minutes - FFS!!!

So, I've decided to keep the CPU GTLs on auto and back down to ~410, I really can't be arsed tuning for the sake of memory performance. Failing that, I have my 400x9@1200 stable profile, although heading even 1MHz upward of this gave me issues. I think 45nm will be the ideal solution, but PSU has priority for now, and hope to possibly have one by next week. After selling my old one along with a handful of other bits, I may be able to stretch to a 9650 in September.

[humeyboy]

Yes obviously in Windows esp Vista with a lot of installed things running the Cores will drift apart and more so the longer it runs but I notice a major difference very soon into the test with Auto over tuned GTL's.

If you actually wanted to make sure all Cores were nearly equal you can run Prime in Safe Mode and they will be near as possible the same as very little Processes running.

[Marcio]

Hi.
This is my 24/7 settings, 30 loops IBT maximum stress level stable, all aircooled, bios 1307.
Priming for 8h so far...

http://img140.imageshack.us/img140/5180/xtremesystem.png

[KURTZ]

SLI ENABLED FOR P45/X38 MAINBOARDS!!!!!

http://en.expreview.com/2009/08/06/exclusive-get-sli-for-your-p45-motherboard.html

[GFORCE100]

SLI ENABLED FOR P45/X38 MAINBOARDS!!!!!

http://en.expreview.com/2009/08/06/exclusive-get-sli-for-your-p45-motherboard.html

Yes, E.V.O. did this on the MIIF too by patching the BIOS but since then I haven't seen him on these forums be it posting in this or other threads. Quite a pity as he also changed some ROM's in the MIIF BIOS.

[mooch49]

will there be a problem running 2 4870's on this board. i've read in various furoms that the p45 has limitations when it comes to crossfire. like it will run the speed at 8x instead of 16x.

[humeyboy]

will there be a problem running 2 4870's on this board. i've read in various furoms that the p45 has limitations when it comes to crossfire. like it will run the speed at 8x instead of 16x.

I can safely say do not listen to anyone telling you that and it was proven to be fine on a few sites many months ago.

The P45 is PCI-E 2.0 and its has 2 x more Bandwidth than previous versions.

I cannot comment on cards soon to be release esp the Dual cards.

[screwtech02]

Your not gonna notice any difference @x8, nuthin comes close to saturating the pci-e bandwidth even @x8....

[KURTZ]

Yes, E.V.O. did this on the MIIF too by patching the BIOS but since then I haven't seen him on these forums be it posting in this or other threads. Quite a pity as he also changed some ROM's in the MIIF BIOS.

i know, but this thing haven't any connection with e.v.o.'s modded bios ... :)

[Mr.Pink]

Try NB GTL's, made major difference from me but same as you above 475MHZ it works 1 reboot and will not next reboot.

Even had 500FSB benching but not for long as it freezes, I blame Bios and lack of Memory support as Corsair ain't even on the list.

Skews did nothing for me, but I cannot boot with a negative NB GTL but can at Auto or Positive.

Currently I am @ +40 NB for 475FSB and CPU GTL's are +20/+40/+20/+40, Tried to leave the 20's at Auto and also tried all Negatives but not sure what's best as of above info. because its simply random each reboot.

It makes no difference if I use 1 Module or all 4 Modules or even my older Corsair Modules, still hard to get over 475FSB and 1175MHZ seems limit though my Ram can do more.

My advise would be to sell of some of the non needed kit and buy a Q9650 and see how far it can go on an A3 after all that's what the Gigabytes are.

that quad still giving you grief???

I found in the beginning positive GTL values worked better for me I'm using a negative value now and am limited to 471FSB above this and my clocks fail...
525FSB is a walk in the park for my 8400 but either my 9550 or 9650 struggles after 470FSB...

[GFORCE100]

i know, but this thing haven't any connection with e.v.o.'s modded bios ... :)

Well yes and no. E.V.O's BIOS patching is a more elegant way of going about it rather than having to install additional software to trick the drivers.

[humeyboy]

that quad still giving you grief???

I found in the beginning positive GTL values worked better for me I'm using a negative value now and am limited to 471FSB above this and my clocks fail...
525FSB is a walk in the park for my 8400 but either my 9550 or 9650 struggles after 470FSB...

Stable for 4 days with multiple reboots now @ 475FSB.

I think its the LLC ain't 100%, I now think I know what to look for in voltages each reboot to know if its going to be stable or not (obv is stable when not Priming).

I need about 1.168v for 3.6GHZ and that means 1.2v is Bios even with LLC as it drops (not droops).

Now it seems to be 1.192v with 1.2v in Bios so LLC is holding better but its not ideal as my previous Mobo which had LLC (NF 780I) would be exactly what you set it to, no VDrop and no VDroop.

I will keep an eye on it every reboot and see if 1.2v ends up 1.168v or 1.192v on load.

I do not get why that would cause issues though as I lowered the Multi and raised Voltages throughout all settings to make up for this to test and it again was a hit and miss.

[Mr.Pink]

I'm currently testing the mosfet temps, I think its the reason this board is so unstable at high FSB..
I have 2 MIIF boards one with stock TIM and cooling the other with mosfet blocks, from just the first few crude tests the water cooled mosfet board clocks higher.

[Haes]

Hi Guys,

bought a Q9550 E0 yesterday. I want to get 4,0 Ghz 24/7 and my first impression is that these little 45nm suckers are really tricky to handle compared to my Q6600 G0.

My Q9550 is a 1,25V VID, pretty standard i would say, but testing Small FFT last night got me turning wild, cause the cpu shows no stable area instead while reducing Vcore stable GTLs swtch over the whole bandwidth.

Small FFT (all BIOS 1901):

8.5*471 (4Gs):

VC/PLL/NB/VTT/CPU GTL

1.35/1.5/1.29/1.2/10 -35 10 -35 stable , else BSOD about 5 to 10 minutes
1.34/1.5/1.29/1.2/60 10 60 10 stable, else BSOD about 5 to 10 minutes
1.33/... nothing stable for now, BSOD about 5 minutes
<1.3/... can boot win7


Got the sucker at 1.36/1.55/1.44/1.31/40 -5 40 -5 stable at 475*8.5(~4.1Gs), else BSOD while stressing

Large FFT BSOD for now

Temps are all normal and 24/7:

4 Gs load 54-58°C
4.1 Gs load 56-60°C

(all air-cooled with Noctua U12P)

...so that's not the problem.

Compared to my Q6600 (http://www.xtremesystems.org/forums/showpost.php?p=3904206&postcount=3797 - could even lower VTT to 1.34V last week) I can't hit the max stable FSB with the Q9550. I need more Vnb to get the same results like with my Q6600 and 475 is for now Small FFT stable but with (for me) no reasonable Volts for 24/7 - even when temps are pretty ok.

Perhaps I've to reset my ocing and take a break.


Anyway, I aprecciate 24/7-settings Q9550 SLB8V with M2F :up: ...

Best Regards

[humeyboy]

I'm currently testing the mosfet temps, I think its the reason this board is so unstable at high FSB..
I have 2 MIIF boards one with stock TIM and cooling the other with mosfet blocks, from just the first few crude tests the water cooled mosfet board clocks higher.

I have a later Mobo (still a Rev A2) but no heat issue with Heatpipes so for now as lazy I left the TIM.

I did ask weeks ago if any7one got more stability redoing the TIM and who also did not have heat issues but got no feedback.

I redid the TIM on the 680I and 7680I as they are hot Mofo's.


@ Haes, IMO no, the 45nm CPU's are easier to clock than 65nm CPU's.

The Q6600 is an old dog and ain't great to get a high FSB on at best of time.

Its just this Mobo to blame TBH, a Dual is a breeze to get high FSB on but a Quad is harder.

The Gigabyte does far better.

[Haes]

I have a later Mobo (still a Rev A2) but no heat issue with Heatpipes so for now as lazy I left the TIM.

I did ask weeks ago if any7one got more stability redoing the TIM and who also did not have heat issues but got no feedback.

I redid the TIM on the 680I and 7680I as they are hot Mofo's.


@ Haes, IMO no, the 45nm CPU's are easier to clock than 65nm CPU's.

The Q6600 is an old dog and ain't great to get a high FSB on at best of time.

Its just this Mobo to blame TBH, a Dual is a breeze to get high FSB on but a Quad is harder.

The Gigabyte does far better.

Well, it can be that Giga does better in the bandwidth with Quads, but with my Q6600 and M2F I get 490FSB stable with reasonable 24/7-setting, temps are really good ~52°C NB at load. So for me Asus did a good job, not brilliant, but good. I agree that a 45nm is better to oc as a 65nm and I hoped to take my settings with few GTL/Skew adjustements to the Q9550. But like we all know each CPU is different, so I've start over.

I forgot to mention, that I tried to stay in a 40-45mV gap like simps guide shows, but I found on forum delux some use 10mV gaps or else with the SLB8V für 24/7 by reasonable volts.

[humeyboy]

No way m8, your going to need put in as much time with the new CPU as you did with the old CPU.

The GTL's are different for a 45nm over a 65nm.

As for the "GAP" in GTL, only the Negatives have 40/45/50 so 0.5's, the Positives have 40/50/60 so 10's

You adjust the GTL going by your VTT (FSB Termination Voltage) into the proper Calculation (in a Sticky in this very section), it could end up nowhere near 40-45, I have seen peeps on +/-60 here

[Haes]

No way m8, your going to need put in as much time with the new CPU as you did with the old CPU.

The GTL's are different for a 45nm over a 65nm.

As for the "GAP" in GTL, only the Negatives have 40/45/50 so 0.5's, the Positives have 40/50/60 so 10's

You adjust the GTL going by your VTT (FSB Termination Voltage) into the proper Calculation (in a Sticky in this very section), it could end up nowhere near 40-45, I have seen peeps on +/-60 here

Yeah, that's what I thought too...

And I resetted all, put all to my starting position... calculated GTL accordingly to the VTT. which would 1.35*0.67-GTL=1.35*0.63 => GTL= 54mV.

So I played up from 40mV with a gap between 40-60mV and get Win7 boot + some benches/validation by 40 -5 40 -5 NB 30 (which is pretty standard for the NB GTL with my M2F) and the CPU GTLs are 1 of 3 functional setups with 471*8.5 by Small FFT that I did last night.

http://www.abload.de/thumb/boot50069550syns.png (http://www.abload.de/image.php?img=boot50069550syns.png)

Vcore is way too much, just don't turned the VC down...

[humeyboy]

I get 1.35v to work out at 0.0432 so 42mV so choice of -40 or -45 or +40 or +50 GTL's

[Haes]

Ok, could bench a little first LargeFFT without errors by 490 FSB like with my Q6600. That's a start :) .

http://img32.imageshack.us/img32/6165/testlargefft4909550.th.png (http://img32.imageshack.us/i/testlargefft4909550.png/)

[seban]

...I have seen peeps on +/-60 here

Well I am +50/+90 nice and stable at 4ghz q9550 ;) - also the higher you go with vtt and vc from nominal values the bigger allowed gap becomes.

And Simps guide is ok for mediocre clocking - if you want to squize max from MB and CPU you got to tune all parameters individually all the time not only at beginning as he suggests - this is why he did not notice that gap is getting bigger as you adjust vcc and vtt while clocking cpu.

[Mr.Pink]

I did not find simps guide very usefull, it did not clock my board higher than my own method....
If I was looking for a 10-15% overclock ..... its built in to the bios

[Haes]

So, I got my Q9550 E0 stable in SmallFFT with 470*8.5 by 1.34V VC , 20/70/20/70 GTL and good temps 55-59°C (Noctua U12P) on the 1901.

It's ok, but my own expectations are a bit higher like 470*8.5 by 1.3-1.32V VC or better 482*8.5 (4.1Ghz) by 1.34V VC.

In fact, I got 475*8.5 (4.05 Ghz) by 20/-25/20/-25 stable in SmallFFT with max 62°C Temps and 1.36V VC.

Some ppl in the forumluxx (german community) have better results with the 1307 Bios,although it wasn't designed for Q9XXX. I'll give it a try and I've E.V.O.'s 1901.2XX Bios too.

A good thing about mine is a very plain templvl. There are times in the stress where all cores have the exact same temp for a long period (1hour or more) and jump together. Never seen such a tempbehavior :) and the max tempdifference is 4°C.

But I've another matter:
My RAMbehavior is really strange. Corsair sent me a new pair (2nd pair already) DDR2 8500er Dominators last week. While with my Q6600 a never had a problem with memtest86+, I've now errors no matter which setting I used all on defaults or loose timings, more VDimm, less frequency or more VNB and no matter which slot or single/dual channel.

Blend is stable by even 460FSB and even ~ 1100MHZ RAM, PL10 or a bit tighter timings and I know Blend isn't such a stress test like Memtest. So what can it be?

Best Regards

[Mr.Pink]

test one stick at a time on stock speeds (1066) , if they pass I would start to play with dram clock skew or up vdimm there good till 2.2volts iirc

[eSp!s0]

Here is an article: "SLI on X38, X48 and P45 possible?": http://www.computerbase.de/news/hardware/mainboards/intel-systeme/2009/august/sli_x38_x48_p45/
The article is in german, so you have to translate it!

[humeyboy]

Corsair are not even on the QVL.

Only the older Rev 1.1 Dominator can do 1200mhz but still cannot do it on this Mobo but can on other Mobos.

[Haes]

test one stick at a time on stock speeds (1066) , if they pass I would start to play with dram clock skew or up vdimm there good till 2.2volts iirc

I did it already and all failed on every slot in default by about 4% to 10% and the curious thing about that is, that with my Q6600 a month ago or so they passed memtest every run.

I do exclude the new cpu, that can't be an error source.
Perhaps something happened with my mobo while ocing, although I never went over the max allowed specifications.

I had to switch my cpu, just to be sure. :(


P.S. Switched back to 2104, set all on default and bumped VDIMM to 2.15V. Memtest seems to run without problems. Very strange indeed, never had any mem issues on 1307 or 1901 at all. Dualchan test is running now.

[humeyboy]

That's the temperamental bug back again.

I changed Bios setting to try higher than 475FSB and after failing loaded my saved profile again with stable 475FSB settings.

Now I'm unstable, this sure is 1 buggy POS.

The fix is to just enter and exit the Bios (don't change anything) till its a stable boot up again.

[Mr.Pink]

Corsair are not even on the QVL.

Only the older Rev 1.1 Dominator can do 1200mhz but still cannot do it on this Mobo but can on other Mobos.

i would say yes they can

http://i237.photobucket.com/albums/ff267/Grnfinger/th_1240ram_2.png (http://s237.photobucket.com/albums/ff267/Grnfinger/?action=view&current=1240ram_2.png) http://i237.photobucket.com/albums/ff267/Grnfinger/th_500x8_1200ram.jpg (http://s237.photobucket.com/albums/ff267/Grnfinger/?action=view&current=500x8_1200ram.jpg) http://i237.photobucket.com/albums/ff267/Grnfinger/th_IMG00015-20090425-1029.jpg (http://s237.photobucket.com/albums/ff267/Grnfinger/?action=view&current=IMG00015-20090425-1029.jpg)

[GFORCE100]

i would say yes they can

http://i237.photobucket.com/albums/ff267/Grnfinger/th_1240ram_2.png (http://s237.photobucket.com/albums/ff267/Grnfinger/?action=view&current=1240ram_2.png) http://i237.photobucket.com/albums/ff267/Grnfinger/th_500x8_1200ram.jpg (http://s237.photobucket.com/albums/ff267/Grnfinger/?action=view&current=500x8_1200ram.jpg) http://i237.photobucket.com/albums/ff267/Grnfinger/th_IMG00015-20090425-1029.jpg (http://s237.photobucket.com/albums/ff267/Grnfinger/?action=view&current=IMG00015-20090425-1029.jpg)

I think what he meant is you cannot do 1200MHz RAM stable on the M2F. I.e. not just claim it's stable with an everest benchmark but do the full agenda which at the minimum would signify Prime95 Large FFT and LinX for a few hours.

Also post your BIOS settings if you're not too lazy :) No seriously, then others can confirm your findings too and put this myth/fact to rest.

[Mr.Pink]

many kits will do 1200MHz on this board, no my 1240 run was not stable
1200 is doable with alot of tweaking, it wont come easy..
My bios profile is here, you just need to look for it

I have a 1200MHz stable shot someplace I'll hunt it down and post it up, iirc it was a prime run for 9 hours so thats good enough for me to call stable
I dont run that Corsair kit 24/7 its a benching kit, I have a PC8500 kit of Mushkins that will do 1200 on this board, sadly my 2 kits of GSkills will not go above 1140-1150MHz

[humeyboy]

Nope I meant what I said.

I and many others inc Gold***** (cant remember his nick) cannot get 1200mhz on ram proved in other Mobo's

I can only get 1175mhz then it will not even post Bios, that's with 1 stick or all 4, no matter what voltage/settings.

Someone here showed Gold**** the 1200mhz Corsair and I think he went back and tried it and finally got 1200mhz but I aint sure if stable

All I know is its not worth the bother, Asus need to get better support in new Bioses.

[Mr.Pink]

I agree the min gains from 1200MHz ram is not worth the hassle...
If Asus had put effort into this board and released a bios that unleashed its full potential I think you would see similar performance that the UD3P has with quads.. sadly this a dual friendly board and seems to hates quads

[humeyboy]

This Memory should post the Bios esp only 1 stick with major voltages, it simply ain't supported.

I'm tempted to get 8GB of 1200mhz G.Skill sent here from USA by a buddy.

I wonder if it will work as its also not supported and if this will help me get 500FSB as it will be o the harder 1:1 ratio.

I stand by the fact we need better Bioses.

[seban]

Giga bords users get higher clocks on quads but check their voltages and temps - those are way higher then on MF2 for same clock - I see giga as benching board and MF2 24/7.

But that is just my opinion.

[Haes]

I'm the same oppinion. M2F for 24/7 and DFI/Giga for Bench. The DFI Lanparty p45 tr2rs plus (ddr3) is my fav P45 Mobo beside the m2f and I still think the m2f is a great looking, super cooling, low volt board thanks to the many phases, just the bios support sucks overall and the bios itself most often too and i don't get used to such a bad service from asus and it's not a rog series board worthy bios support.


Got a dominator stick 5-5-5-15-3-55 with 1120 by 1.99VDIMM PL10 moderate Memtest stable.
Testing the others now.

[In-Fluence]

Here are my settings for getting the CPU stable at 400x9:


Processor: Q6600
Ai Overclock Tuner: Manual
OC From CPU Level Up: Auto
Ratio CMOS Setting: 9
FSB Frequency: 400
CPU Clock Skew: Normal
NB Clock Skew: 400
FSB Strap to North Bridge: 266
DRAM Frequency: DDR2-1200MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 15 DRAM Clocks
RAS# to RAS# Delay: 3
Row Refresh Recycle Time: 50 DRAM Clocks
Write Recovery Time: 5
Read to Precharge Time: 3

Read to Write Delay (S/D): 8
Write to Read Delay (S): 3
Write to Read Delay (D): 5
Read to Read Delay (S): 4
Read to Read Delay (D): 6
Write to Write Delay (S): 4
Write to Write Delay (D): 6

Write to PRE Delay: 11
Read to PRE Delay: 5
PRE to PRE Delay: 1
All PRE to ACT Delay: 5
All PRE to REF Delay: 5
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Disabled
Ai Clock Twister: Normal
Ai Transaction Booster: Manual

Common Performance Level: 6
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled

PCIE Frequency: 100

CPU Voltage: 1.4375
CPU PLL Voltage: 1.553
FSB Termination Voltage: 1.312
DRAM Voltage: 2.15775
North Bridge Voltage: 1.40475
South Bridge 1.5 Voltage: 1.5265
South Bridge 1.1 Voltage: 1.15300

CPU GTL Reference (0): +50
CPU GTL Reference (1): +10
CPU GTL Reference (2): +50
CPU GTL Reference (3): +10
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 9
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled

Load-Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

I found that adjusting the GTLs to +50,+10 gave me 9 hours small fft stability (which I stopped), whilst anything lower or higher conked out after 8 minutes or less. This was also tested twice on each setting, so I'm now much more confident this is the correct one for me.

I've now pushed up to 405x9 which is pushing my memory a little more, but I know it's stable up to 1233 so hopefully I can begin to encroach on my old 410x9 on 1.4625v setting I had on the P5B.. At least I'm close, with faster memory speeds :)
So far 405x9 failed on the 4th core (which I know is weakest) with the same vcore, so am currently testing with 1.45v which has improved so far.

410x9 is still crapping out on me even at 1.5v so I would ask for any suggestions on how to approach this speed.

[WaterFlex]

2202 Bios is available (ftp://ftp.asus.com.tw/pub/ASUS/mb/socket775/Maximus_II_Formula/Maximus-II-ASUS-Formula-2202.zip)

[Haes]

Flex was faster :) ...

Still tweaking with the Q9550 but it gets better and better.

[humeyboy]

WOW, What have Asus f00ked now. :p:

Edit : Seems they have f00ked their site, freezes at drop down choices for me, used FTP instead.

Edit 2 : Finally got on the Asus site, Bios is not listed, only on FTP so no info on changed but I guess the usual "Added Memory Support BS", still locks up at Bios if I try 1200mhz.

[In-Fluence]

Testing it now, seems to be as stable as 1307 with the same settings above, and could post at 490 which I couldn't before. It still only shows 2 cores, and I could have got lucky with gtls or something.

On the site the changelog says the same as the others: 'improve memory compatibility'

One thing I did notice with 1307 is that it reports the nuber of cores on post but 2104 and 2204 don't.

Thanks for the heads up waterflex:up:

[Haes]

[ 2202 ]
----------------------
Fix CPU Fan speed wrong after resume from S3


That's it !

Will install on the 2nd Bios chip later and test it.

[In-Fluence]

Oops. my bad, misread that - it's obviously just appeared must have read the 2104 :s

Where you read that from?

p.s - anyone care to offer any further advice re: the above settings?

[Haes]

I don't have the link from the original source, but I copied it from the Maximus 2 Formula Thread [from the Hardwareluxx Community/Germany], because I asked about changes there and somebody copied that from Asus' hp without linking it.

P.S. Influence, can you post your hole 490 fsb setting here ? Will see, even if I don't installed my G0 now, but I saved my results and settings.

P.S.S. Tried the new bios with my 4.0 Ghz stable settings witz 2104. I had BSOD all the way down, even bumping Vcore one step up was useless. Switched immediately back.

[Mr.Pink]

Here are my settings for getting the CPU stable at 400x9:


Processor: Q6600
Ai Overclock Tuner: Manual
OC From CPU Level Up: Auto
Ratio CMOS Setting: 9
FSB Frequency: 400
CPU Clock Skew: Auto
NB Clock Skew: Auto
FSB Strap to North Bridge: 266
DRAM Frequency: DDR2-1200MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 15 DRAM Clocks
RAS# to RAS# Delay: 3
Row Refresh Recycle Time: 50 DRAM Clocks
Write Recovery Time: 5
Read to Precharge Time: 3

Read to Write Delay (S/D): 8
Write to Read Delay (S): 3
Write to Read Delay (D): 5
Read to Read Delay (S): 4
Read to Read Delay (D): 6
Write to Write Delay (S): 4
Write to Write Delay (D): 6

Write to PRE Delay: 11
Read to PRE Delay: 5
PRE to PRE Delay: 1
All PRE to ACT Delay: 5
All PRE to REF Delay: 5
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Disabled
Ai Clock Twister: Normal
Ai Transaction Booster: Manual

Common Performance Level: 6
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled

PCIE Frequency: 100

CPU Voltage: 1.4375
CPU PLL Voltage: 1.553
FSB Termination Voltage: 1.312
DRAM Voltage: 2.15775
North Bridge Voltage: 1.40475
South Bridge 1.5 Voltage: 1.5265
South Bridge 1.1 Voltage: 1.15300

CPU GTL Reference (0): +50
CPU GTL Reference (1): +10
CPU GTL Reference (2): +50
CPU GTL Reference (3): +10
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 9
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled

Load-Line Calibration: Enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

I found that adjusting the GTLs to +50,+10 gave me 9 hours small fft stability (which I stopped), whilst anything lower or higher conked out after 8 minutes or less. This was also tested twice on each setting, so I'm now much more confident this is the correct one for me.

I've now pushed up to 405x9 which is pushing my memory a little more, but I know it's stable up to 1233 so hopefully I can begin to encroach on my old 410x9 on 1.4625v setting I had on the P5B.. At least I'm close, with faster memory speeds :)
So far 405x9 failed on the 4th core (which I know is weakest) with the same vcore, so am currently testing with 1.45v which has improved so far.

410x9 is still crapping out on me even at 1.5v so I would ask for any suggestions on how to approach this speed.

try raising your PLL a bump or 2
FSBT might have to be increased also but I would start with PLL

CPU Clock Skew will also help you alot , ther is no formula for it you'll have to test it untill you find the sweet spot

[humeyboy]

I think if my Mobo was an A3 I could get 500FSB stable, cause I can bench from 475-500FSB but ain't stable and totally random with settings.

Would seem 475FSB is my max stable (can change if I mess with settings then change back again, got to leave alone once stable).

I wonder if Asus will make 1 more Mobo to compete with this new 24 phase Gigabyte that's coming, I ain't going to 1366 for a while yet.

There is some new kits of DDR2 1200 just out so its got life in it yet.

[Haes]

I would recommand the same procedure... try with pll or 1 or 2 steps up with the vtt, but first I would try the pll and also try to set the cpu / nb skew.

I have 490 FSB custom and large stable with 24/7 settings.
But set the multi to the max would give me about 4165 Mhz and that's too much for my Q9550 (max is about 4120) with 24/7 setup.
I set the hole system by 4,04 Ghz (471*8.5) and try to lower the volts as much as possible and there I made same progress and will test a bit.

If an A3 can do better as my A2 ? Well, I don't know, some perhaps but there are also some which perform worst (like cpu steppings/vids)
So, I'm satisfied.

Didn't know, that Giga build an 24 phases mobo.

@ Humeyboy : Very good DDR2 9600Kit is the OCZ Blade Series (http://www.ocztechnology.com/products/memory/ocz_pc2_9600_low_voltage_blade_series) (see waterflex).
I'm considering myself to buy a kit to use stable 1130+ and you can buy them in black pcb too.

[In-Fluence]

I tried adjusting both voltages but anywhere between 401-406 FSB always resets on 12kFFT test - this has been my main issue from the start. I'm certain overheating isn't the issue as I have a household fan pointing staright into the side of the case. Aside from more pll, vtt (while re-testing gtls) and even adding a little nb I'm at a bit of a loss :/
The only thing I can think of that could be the cause for this now is when I raise the vtt to above 1.31v, the GTL gap needs 45mv but is best suited to positive GTL values meaning I can only tune to within 40mv (and have tried all the -ve values with similar or worse results). The CPU liked 1.2vtt on the P5B, but once again this just made it less stable on here.

I'm hating myself for wasting time as much as you lot probably are right now, but I've gone back to my 457x8 with 1141 memory profile, as this gives me the higher fsb and memory bandwidth and it seems a lot easier to attain higher overall clocks. It just seems strange to me that 410x9 is so hard to acheive.

I have already determined the best skew values at normal/400. 100/300 worked nearly as well, but not at higher fsb values. All other combinations resulted in much less stability, and all combinations were tested at 400 and 457 fsb. To get near to the 470+ range 100/600 - 200/700 was required. (Sorry i didn't fill those values out they should read normal/400)!

2204 turned out to be a load of crap for me, so I headed back to trusty old 1307. I might consider 1901 once I establish my final overclock for 24/7.

[DaLiu]

I manage 475 fsb full stable for my Q9450 (475 x 8), the ram is on 1140 Mhz with 1,38 on NB but i still have the same temp problem with the :banana::banana::banana::banana:ty NB, right now idle is 57 and full load 66, i tried 3 times to fix it with washers and changing the paste, nothing helped, the mobo is bended and the contact between chipset and heatsink is not very good, at least looks like even running IBT or games for hours the temp on NB is not going higher than 66.

http://lab501.ro/forum/attachment.php?attachmentid=750&d=1249807698
http://lab501.ro/forum/attachment.php?attachmentid=751&d=1249807698

[Mr.Pink]

a guy on OC forum runs gtls +50 for all 4 and NB gtl at +20 for 471 FSB on a Q9550, 9 hours prime stable... something to consider trying

I have seen a couple of A3 boards they dont do any better , I think the MIIF is just a dual cpu board and it hates quads, to bad Asus wont releases a better bios..
Maybe its time to try a UD3P

[humeyboy]

Another waste of time Bios.

Though I had 480FSB with 1 more notch or Voltage to NB+VTT.

7 hours looking good but it froze overnight.

Ok up voltages 1 more notch (really low anyhow under 1.32), locks up in 30mins, as most will now more voltage on a P45 is not always a good thing.

Not worth the BS for an extra 5MHZ FSB never mind trying to get to 500FSB.

[humeyboy]

@ Haes, OCZ (never liked them period) used to have major issue with Asus as they claimed Asus used a non Ref Bios (NF680I and more).

If I bought any Memory it would be the new 1.8v 1200mhz 4GB (x2) G.Skill Trident, I can already run all 8GB (same as 1GB) at 1175MHZ with no fancy tweaks.

Problems is neither of that Memory is on the QVL cause Asus are lazy bastards.

[In-Fluence]

http://i194.photobucket.com/albums/z182/in-fluence/365GHzn1219testing.jpg

Here's what I've managed and will be intent on getting 100% 24/7 stable. This hung after about 3 hours of blending on ~384KFFT length, but passed the 12k test. CPU is still failing at around 12K on the small test though. VTT is only at 1.3v though so I'm thinking a bump to around 1.32-1.35 might help this, would you agree?

Running GTLs at all +50 seemed to even out the timing of all the tests but it restarted at around the 3rd round of 12K tests. Currently testing at 50,10,50,10 to see what the difference is, then will bump up the vtt. If I do this, how should go about adjusting the GTLs accordingly? Would an increase in vtt normally require an increase in GTL refs on 65nm, or is there no certain pattern?

[Haes]

Wow, 1,48V VCore is a lot more than my old g0 needed (was about 1.43V), but nethertheless okay. I would try with vtt if cpu clock didn't worked or gtls and use your last stable gtls/clocks and just bump vtt. if this not worked try gtls than clocks... and so on (this ist my system and work for me very well).

GTLs calculation accordingly to Intel Specs:

VTTx0.67=GTL [mV] in general.

Gap calculation:

C2D: VTTx0.67-VTTx0.63= +/- GTL Gap
C2Q: VTTx0.7-VTTx0.67= +/- GTL Gap

Functional GTL range are between +/- 5% of VTTx0.67 for C2Q and ...x0.63 for C2D.

For example C2Q:

VTT needed 1.265V, so...

1.265V x 0.67 = ~ +/-85mV (90mV)

gap: 1.265V x 0.7 - 1.265 x 0.67 = +/-38mV (40mV)

So:

90/50/90/50 or 50/90/50/90 or -90/-50/-90/-50 and so on and from there +/- 5%. For some it works very well, for others not.


But it's not 100% right, cause every combination of mobo,cpu,ram,temp and bios ist different.

@ humeyboy: thx for info

[Eeky NoX]

Gat my old modules of D9 (gmh) on UD3 now, 'n gat 740mhz now !! Only 666 with the same 2,6v on the MIIF ^^ ...but it was better to get the E8600 @4,8 stable on air :D

[humeyboy]

Only run 2GB today ?, hope its not Vista.

[Mr.Pink]

Here's what I've managed and will be intent on getting 100% 24/7 stable. This hung after about 3 hours of blending on ~384KFFT length, but passed the 12k test. CPU is still failing at around 12K on the small test though. VTT is only at 1.3v though so I'm thinking a bump to around 1.32-1.35 might help this, would you agree?

Running GTLs at all +50 seemed to even out the timing of all the tests but it restarted at around the 3rd round of 12K tests. Currently testing at 50,10,50,10 to see what the difference is, then will bump up the vtt. If I do this, how should go about adjusting the GTLs accordingly? Would an increase in vtt normally require an increase in GTL refs on 65nm, or is there no certain pattern?

up your VTT its should pass the 384FFT test

[KURTZ]

hi folks, news about the 2202? anyone have tested it?

[humeyboy]

hi folks, news about the 2202? anyone have tested it?

LOl, I just ranted about it above.

Same old crap as previous ones, ain't even on the Asus website.

[KURTZ]

OK :D

just a Q humeboy: so the A3 revisions aren't good mobos? the better are still the A2?

[humeyboy]

OK :D

just a Q humeboy: so the A3 revisions aren't good mobos? the better are still the A2?

No, I have an A2, would be better with an A3 as its newer and the good clocking Gigabytes are A3's.

Someone here had 490FSB on a Q6600 on one, I cannot get that stable on a Q9650.

[Mr.Pink]

After some testing 1307 is ( for me ) the best bios for quads

[humeyboy]

Tried it too, exact same results.

This Mobo is a POS for Quads IMO.

[KURTZ]

After some testing 1307 is ( for me ) the best bios for quads

1307 is still the best bios, imho :)

[Haes]

No, I have an A2, would be better with an A3 as its newer and the good clocking Gigabytes are A3's.

Someone here had 490FSB on a Q6600 on one, I cannot get that stable on a Q9650.

That's me and I got it with the same /VTT/SB/NB GTL etc. stable with my 9550 E0, but setting the multi high it get unstable and I've to bump NB to 1.4V (Q6600 was 1,36V NB - very strange), have to tweak the VCore CPU GTLs, PLL and VTT.

My problem is sometimes with the same settings it 2-3 hrs smallfft stable (max multi) and sometimes not even 15 mins.

And tweaking the GTLs.VTT,PLL is for 4,0+ Ghz +/- 50 Mhz is really hard with my E0.

[humeyboy]

That's why for a guaranteed 4GHZ on this Mobo I advise a Q9650, all are E0 so no luck of the draw and the higher Multi.

Once you have a stable FSB it should not be affected by higher MHZ on the Multi.

The only things that you should need adjust is VCore and PLL.

I have same issues as you, 1 setting is stable for hours then fails, then its only stable for a few mins, then its again 100% stable.

[Haes]

that's what i thought too and this was the reason hy i tweaked out the max stable fsb with mymobo to be 100% sure that i can run 4gs (471*8,5) with a 9550 (if the 9550 will do it) and i got results with small fft stable with 1,34V VCore.

Now you add stable 490 fsb settings to the stable vcore settings and you expect stable 24/7 4g 9550 rig but no ...

Even testing different bios versions keep me back to the most often used versions:

1307
1901
2104

and with no improve.

[KURTZ]

just wondering guys, if i put a second nVdia card for the physics, the first pci slot goes @ 8x?

[Haes]

no matter which pcie slot...

if u use only one card you've a 16x lane connection and if you use CF you've on each lane 8x .

[humeyboy]

Don't matter its PCI-E 2.0, so 8x is same as old 16x.

[In-Fluence]

Just for the record, if you use just the lower PCI-E slot in its own it goes to 8x too. I still need to mount my HR-03 the other way up so it fits in there :)

Still having inconsistent results at 457x8 - on certain tests it will pass 12K and continue to do it for 3 entire small fft loops before I stop it. I then reboot and it fails a minute or 2 after it hits 12k - grrrr!

[Haes]

I think seriously my M2F have n memoryslot problem with 1901 - have to check it again with memtest and switch to 2104, cause there i passed with all four slots 1066mhz.


I got rounding errors after only three minutes (core 3 and 4) custom inplace run 4096 ffts 15minutes each fft with 920mhz with 2.1V and than i put only the one memory in the slot where i know that it makes 1120 with 1.99V.

And know custom runs 1hr and no problem at all :rolleyes: ... , besides I set 1,42V Vcore LLC off VCore auto and tested with 460*8 and VCore drops to 1,32V stable ... wtf 0.08-0.1V drop that's huge :shocked: ...

[humeyboy]

Yip even with LLC there is a drop/droop on mines.

On my old 780I it was solid with LLC on.

[Mr.Pink]

strange with LLC enabled my vcore actually goes up under load

[Haes]

If Vcore goes up and under , it means you have to bump your gts 1 or 2 notch.
It's stable but not 100%... this is why you've GTLs ... to adjust correctly the volts.

[In-Fluence]

strange with LLC enabled my vcore actually goes up under load

Same as me, as much as .025v which takes me to 1.5v, and things are getting a little toasty in here!

I've managed to get a bit further with 60,20 gtls, but it hung within a minute of the 14k test. Strange as with 50,10 (if it did get past 12k), it always got past 14K and higher ffts. Currently testing 70,30 to see if it gets further, but it seems to have raised temps a fair bit :(

[Haes]

Well, I will let my Q9550 something in between 3.6-3.9 Ghz, cause that stupid thing is really bad or I was unlucky.

I got even 4Ghz with reasonable volts stable with custom inplace run for about 3 hrs. Would I play a little bit I could perhaps make this rockstable but I tweaked over 1 week already and need a break.

If I set the multi on 6 and try my 490FSB setup it's rockstable with a little bump of 1 or 2 notches on the NB (Q6600 setup needs less).

But the cpu demands volts volts and again volts.
1.34-1.37 ~4Ghz, 1.31-1.33 ~3.9 Ghz, 1.29-1.31 ~3.8Ghz. Really thought about selling it and buy a Q9650.

I could even boot win7 with 4 Core 505FSB. Even with 1.41V by 4.1Ghz the cpu is 66°C under load small fft (stable vor 1 hrs or so), so cooling and mobo isn't the issue here it's just the cpu.

[In-Fluence]

Anyone know of a good vcore read point for a DMM measurement? (Ideally for all voltages would be great). I'm tempted to stick one in there & try & find it myself, but as I'm sure you're all aware, it's a risky business, and only have a rough idea where to look! :)

p.s. 50,10 is my most stable GTL setting with 1.31 vtt and 1.40nb at both 450x8 and 400x9 so I have a good reference point to go from.

A slight increase to 1.32vtt helped me stabilise a fair bit more at 450x8 with 1200@5-5-4-5-15 PL7 (1 core failed after 38mins during 768K test, 2nd core failed in 1.5 hours during 448k test, the rest kept going for over 3 hours).
1.31vtt only got me 10 mins at the same settings, so I now feel the urge to up the VTT a bit more, but would like to know if anyone has advice on how the GTLs would need to be adjusted accordingly.
Setting any higher vtt with current GTLs only gives me 10 mins of stability again, plus the 45mv gap comes into play after 1.32vtt (according to the vtt/GTL formula) so I'm suspecting the GTLs will need to be re-tuned - hurrumph! :/

*edit:
Backing the PL down to 9 and setting NBGTL to -10 got me 100% stable albeit at the cost of memory bandwidth, but still better than 400x9, and I at least know what's causing the instability now so I can hopefully tweak further :)
http://i194.photobucket.com/albums/z182/in-fluence/450x8-36GHz1200Stable.jpg

[64dragon]

hey guys, i'm having some issues clocking my new chip (but all other hardware is the same), it's a Q9505 (9550 but with 6mb cache). i'm at 3.5ghz (412*8.5)@1.306v in bios, vNB and VTT at 1.33v, i'm 9hrs orthos small FFT stable and 10hrs large FFT stable BUT when i start running WCG i get computational errors and BSOD. ram is underclocked @ 989 and 2.18v (same voltage i've been running since October with my X3330). With the X3330 i needed 1.4v for NB and VTT. Any ideas what i need to adjust?

[WaterFlex]

2202 is ok for me.

[In-Fluence]

@64 Dragon - Have you tuned your CPU and NB GTLs? What are your temperatures like, and do any of them get higher in WCG than when testing? I only mention this because I've recently discovered that my PWMs get too hot and have the same kind of effect.

On another note, I've been running a little experiment to determine the effectiveness of the Asus cooling solution to my own consisting a NB waterblock on the P45's die (Shared with CPU), 2x copper enzotech mosfet sinks and a aluminium chipset HS for the SB. Here's what I recorded:

______|1___|2___|3___|4___
CPU__|72°C |71°C|82°C |78°C
NB___|47°C |51°C|47°C |51°C
SB___|40°C |48°C|40°C |49°C
MB___|34°C |34°C|34°C |34°C
PWM_|60°C |57°C|64°C |61°C

1: Custom cooling - 1024K FFT test (31°C ambiemt)
2: Asus cooling - 1024K FFT test (28°C ambiemt)
3: Custom Cooling 8K FFT test (31°C ambiemt)
4: Asus Cooling - 8K FFT test (28°C Ambiemt)

I set the motherboard to all auto, with CPU level up to 'Crazy' giving 389*9 with memory at 933. Voltages were left to generate the heat, and were as follows:
Vcore: 1.488
Mch: 1.32
VTT: 1.31
PLL: 1.79
SB/1.1: 1.5/1.12
DDR: 1.86

As you can immediately see, the SB benefits greatly from being cooled on its own by plummetting by nearly 10°C, and that's with 3°C Higher ambient temps! It's clear from this alone that a lot of heat spreads towards this part of the Asus cooler
The northbridge shows expected results with a drop of 4°C (probably more if the ambiemt temps were the same) when it is cooled on its own.
The PWM seems to suffer the most when cooled on its own as it runs 3°C higher, and although this could be down to the ambiemt temps, judging by the part of the case it is in (i.e. v.little airflow in that area) I would say that it needs the extra cooling when running with higher voltages.

So it looks like I either need some mosfet blocks, a better PSU (a small chance), or will have to get another fan going on the mosfet area (which I really want to avoid, as I'm aiming for a semi-passive setup). Does anyone experience worse PWM temps with a naff PSU compared to a good one, or can anyone test this for me? This will probably make my decision to go for a Corsair HX850 at the end of this week, so accurate feedback would be very useful in making a decision about this issue.

[64dragon]

@64 Dragon - Have you tuned your CPU and NB GTLs? What are your temperatures like, and do any of them get higher in WCG than when testing? I only mention this because I've recently discovered that my PWMs get too hot and have the same kind of effect.

In Orthos I see max cpu temp of 51C and NB and SB are 40-45C and i've not seen the PWM break 40C. WCG gives lower temps than Orthos

no, i've not tuned the GTLs, i'm only at 412FSB and people say you can go to 450 before needing to tune them. On my Xeon i was at 438FSB and didn't tune them so i figured a lower FSB would be fine. Guess i'll have to learn how.

[Mr.Pink]

hey guys, i'm having some issues clocking my new chip (but all other hardware is the same), it's a Q9505 (9550 but with 6mb cache). i'm at 3.5ghz (412*8.5)@1.306v in bios, vNB and VTT at 1.33v, i'm 9hrs orthos small FFT stable and 10hrs large FFT stable BUT when i start running WCG i get computational errors and BSOD. ram is underclocked @ 989 and 2.18v (same voltage i've been running since October with my X3330). With the X3330 i needed 1.4v for NB and VTT. Any ideas what i need to adjust?

Post your bios setup

You have tested the ram with memtest to rule out bad dimms??

[64dragon]

Post your bios setup

You have tested the ram with memtest to rule out bad dimms??

i did one pass last night and it was fine (i know i should do more), i've been running this ram in this board since October and the ram has never even hit the rated 1066. while i was at work today i ran 8hr of OCCT large just fine

Processor: Q9505
BIOS revision: 1802
Ai Overclock Tuner: Manual
Ratio CMOS Setting: 8.5
FSB Frequency: 412
CPU Clock Skew: auto
NB Clock Skew: auto
FSB Strap to North Bridge: 333
DRAM Frequency: DDR2-990MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 18 DRAM Clocks
RAS# to RAS# Delay: 3
Row Refresh Recycle Time: 55 DRAM Clocks
Write Recovery Time: 6
Read to Precharge Time: 3

Read to Write Delay (S/D): 8
Write to Read Delay (S): 3
Write to Read Delay (D): 5
Read to Read Delay (S): 4
Read to Read Delay (D): 6
Write to Write Delay (S): 4
Write to Write Delay (D): 6

Write to PRE Delay: 14
Read to PRE Delay: 5
PRE to PRE Delay: 1
All PRE to ACT Delay: 6
All PRE to REF Delay: 6
DRAM Static Read Control: auto
DRAM Read Training: auto
MEM. OC Charger: auto
Ai Clock Twister: Strong
Ai Transaction Booster: Manual

Common Performance Level: 12
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
...
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled
...

PCIE Frequency: 101

CPU Voltage: 1.30625
CPU PLL Voltage: 1.5265
FSB Termination Voltage: 1.3385
DRAM Voltage: 2.118
North Bridge Voltage: 1.3385 (also tried 1.3517
South Bridge 1.5 Voltage: 1.51325
South Bridge 1.1 Voltage: 1.11325

CPU GTL Reference (0): auto
CPU GTL Reference (1): auto
CPU GTL Reference (2): auto
CPU GTL Reference (3): auto
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 8.5
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled

Load-Line Calibration: Disabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

[seban]

Why you have performance level set to 12? You should be able to go 8 without a problem.

Also I would try adjusting NB GTL and / or CPU and NB clock skew.

You should easily see the difference between settings and stability and those might fix your problems.

[64dragon]

i'll worry about the performance level when i'm done OCing. how do i know when i need more PLL? i upped that a notch when i was in bios getting those settings and upon boot it ran WCG for a little over an hr fine, just shut it down to modify something on my case (on the laptop now).

[In-Fluence]

Bios 1802? Bleurgh! Use 1307 or 1901 to rule out some of the issues others have had with the not-so good bios. 1307 has worked best for me. I found that an increase in PLL was needed to get me stable but haven't found the need to go beyon 1.55v (yet) - normally for 450+ FSB speeds imo, but I would say it's worth trying a little more to see if that helps. Pll voltage is pretty harmless until you go beyond the 1.6v mark.

This portion of your settings might need a little adjustment:
Write to Write Delay (D): 7

Write to PRE Delay: 14
Read to PRE Delay: 5
PRE to PRE Delay: 1
All PRE to ACT Delay: 6
All PRE to REF Delay: 6
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: Enabled
Ai Clock Twister: Moderate

Sorry, had to dash off mid-post - your settings do actually look pretty relaexed considering, so the above settings probably wouldn't help a great deal, but I found the Mem OC charger setting to help me, even at lower speeds. CPU & NB skews are definitely the ones to look out for. Bigger gaps between the CPU & NB Skew seemed to offer me loads more stability, particularly at FSB higher than 360. I use normal/400, but it's probably a lot different for your 45nm.

[Mr.Pink]

i did one pass last night and it was fine (i know i should do more), i've been running this ram in this board since October and the ram has never even hit the rated 1066. while i was at work today i ran 8hr of OCCT large just fine

Processor: Q9505
BIOS revision: 1802
Ai Overclock Tuner: Manual
Ratio CMOS Setting: 8.0
FSB Frequency: 450
CPU Clock Skew: Normal
NB Clock Skew: Normal
FSB Strap to North Bridge: 333
DRAM Frequency: DDR2-1081MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 15 DRAM Clocks
RAS# to RAS# Delay: 3
Row Refresh Recycle Time: 55 DRAM Clocks
Write Recovery Time: 6
Read to Precharge Time: 3

Read to Write Delay (S/D): 8
Write to Read Delay (S): 3
Write to Read Delay (D): 5
Read to Read Delay (S): 4
Read to Read Delay (D): 6
Write to Write Delay (S): 4
Write to Write Delay (D): 6

Write to PRE Delay: 14
Read to PRE Delay: 5
PRE to PRE Delay: 1
All PRE to ACT Delay: 5
All PRE to REF Delay: 5
DRAM Static Read Control: Enabled
DRAM Read Training: Enabled
MEM. OC Charger: Disabled
Ai Clock Twister: Stronger
Ai Transaction Booster: Manual

Common Performance Level: 8
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
...
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled
...

PCIE Frequency: 101

CPU Voltage: 1.30625
CPU PLL Voltage: 1.5265
FSB Termination Voltage: 1.3385
DRAM Voltage: 2.118
North Bridge Voltage: 1.39150
South Bridge 1.5 Voltage: 1.51325
South Bridge 1.1 Voltage: 1.11325

CPU GTL Reference (0): auto
CPU GTL Reference (1): auto
CPU GTL Reference (2): auto
CPU GTL Reference (3): auto
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 8.0
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled

Load-Line Calibration: Disabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

Suggestios are in bold for a 3.6GHz clock with decent bandwidth, some volts will have to be adjusted and GTL's will be the key to stability on this board.... Just noticed your run LLC disabled try to enable it and retest you might be surprised on the reults

This is my Q9550 setup

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 8.5
FSB Frequency : 450
CPU Clock Skew : Normal
North Bridge Clock Skew : Normal
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-1081
Dram Clock Skew CH1 A1 : Normal
DRAM Clock Skew CH1 A2 : Normal
Dram Clock Skew CH1 B1 : Normal
Dram Clock Skew CH1 B2 : Normal


DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time : 55
Write Recovery Time : 6
Read to Precharge Time : 3

Read to Write Delay (S/D) : 8
Write to Read Delay (S) : 3
Write to Read Delay (D) : 5
Read to Read Delay (S) : 4
Read to Read Delay (D) : 6
Write to Write Delay (S) : 4
Write to Write Delay (D) : 6

Write to PRE Delay : 14
Read to PRE Delay : 5
PRE to PRE Delay : 1
ALL PRE to ACT Delay : 5
ALL PRE to REF Delay : 5

DRAM Static Read Control: Enabled
Dram Read Training : Enabled
MEM OC Charger : Disabled
Ai Clock Twister : Stronger
Transaction Booster : Manual

Common Performance Level [8]

Pull-In of CHA PH1 Disabled
Pull-In of CHA PH2 Disabled
Pull-In of CHA PH3 Disabled
Pull-In of CHA PH4 Disabled
Pull-In of CHA PH5 Disabled
Pull-In of CHB PH1 Disabled
Pull-In of CHB PH2 Disabled
Pull-In of CHB PH3 Disabled
Pull-In of CHB PH4 Disabled
Pull-In of CHB PH5 Disabled

PCIE Frequency : 101

CPU Voltage : Whatever your CPU needs
CPU PLL Voltage : 1.50
FSBT : 1.33850
DRAM Voltage : 2.12
North Bridge Voltage : 1.39150
South Bridge Voltage 1.5 : 1.5
South Bridge Voltage 1.1 : 1.1

CPU GTL Reference 0 : +20mv
CPU GTL Reference 1 : +10mv
CPU GTL Reference 2 : +10mv
CPU GTL Reference 3 : +20mv
North Bridge GTL Reference : AUTO
DDR2 Channel A REF Voltage : AUTO
DDR2 Channel B REF Voltage : AUTO
North Bridge DDR Reference : AUTO

Load Line Calabration : Enabled
CPU Sread Spectrum : Disabled
PCIE Spread Spectrum : Disabled


http://i237.photobucket.com/albums/ff267/Grnfinger/th_3900_stable.jpg (http://s237.photobucket.com/albums/ff267/Grnfinger/?action=view&current=3900_stable.jpg)

[64dragon]

PLL at 1.55 didn't help, woke up to a BSOD saying something about a clock interupt on a seconday processor
a quick search brought this up but i can't see it at work, it's blocked http://forums.extremeoverclocking.com/showthread.php?p=3205789#post3205789

i'll try your suggestions tonight and see what happens

[lha]

Hi!
I have this MIIF mobo and in my opinion its really great P45 mobo. It doesnt reach like 600-700 FSB but all below that are ok to me :D
I have now E8500 processor @ stock or sometimes @ 4Ghz and a palit GTX 295. So if i buy Q9650 will my "bottleneck" from graphic card disappear? I know the Q9550 is lot more cheaper but it has lower multi and i think that if i buy a Quad processor then i buy little better and expencive.

[In-Fluence]

64Dragon - Try finding your CPU & NB Skews, they're quite a lot easier than the GTLs. In most settings I've seen posted in this thread, the NB is normally delayed by 100-300ps more than the CPU (although mine prefers normal/400 which I haven't seen much). If you try setting the highest fsb that you can boot from and get a short period of prime blend before failing, then try setting CPU & NB skews to normal/normal, then normal/100, normal,200....100/100,100/200 etc, making note of stability times, one or two settings should stand out a lot more
Some won't boot at all or even post, others will give you 5-10 ins stability, and 1 or 2 may give you a lot more, perhaps even completey stabilise it! This is how it worked for me anyway. GTLs are a PITA to find though, but do help a lot once you've found a sweet spot.

Hi LHA, not sure about that - it makes no real difference on my 8600GTS with an E6600 or a Q6600, but that's a tad puny in comparison! It might depend on the game, but I rekon the bottleneck would only really start to show on SLI/Xfire setups. I'm sure someone else here could answer you better :)

[Mr.Pink]

64Dragon - Try finding your CPU & NB Skews, they're quite a lot easier than the GTLs. In most settings I've seen posted in this thread, the NB is normally delayed by 100-300ps more than the CPU (although mine prefers normal/400 which I haven't seen much). If you try setting the highest fsb that you can boot from and get a short period of prime blend before failing, then try setting CPU & NB skews to normal/normal, then normal/100, normal,200....100/100,100/200 etc, making note of stability times, one or two settings should stand out a lot more
Some won't boot at all or even post, others will give you 5-10 ins stability, and 1 or 2 may give you a lot more, perhaps even completey stabilise it! This is how it worked for me anyway. GTLs are a PITA to find though, but do help a lot once you've found a sweet spot.

Hi LHA, not sure about that - it makes no real difference on my 8600GTS with an E6600 or a Q6600, but that's a tad puny in comparison! It might depend on the game, but I rekon the bottleneck would only really start to show on SLI/Xfire setups. I'm sure someone else here could answer you better :)

His FSB should not require NB or CPU skews there benifical for higher FSB
GTL's are a simple math equation and will benifit him greatly, this boards key to stability is GTL's

[In-Fluence]

That's why I said it is what works for me, and I'm simply offering a suggestion for him. FYI, my system is not even stable for 10 mins at 400FSB and I can't even post at 450 without the skews being tuned. If I have the correct GTLs tuned and not the skews, I still can't get near 450FSB, so from my experience, the skews have a more profound effect than the GTLs, I'm not saying it's written in stone for every user! The GTL equation certainly isn't either!

[Mr.Pink]

That's why I said it is what works for me, and I'm simply offering a suggestion for him. FYI, my system is not even stable for 10 mins at 400FSB and I can't even post at 450 without the skews being tuned. If I have the correct GTLs tuned and not the skews, I still can't get near 450FSB, so from my experience, the skews have a more profound effect than the GTLs, I'm not saying it's written in stone for every user! The GTL equation certainly isn't either!

No need to get offended, my comment is based on almost a year of testing.... There was nothing personal intended
I have 2 MIIF boards skews have little impact untill I go above 480FSB

[64dragon]

Mr.Pink and In-Fluence, thank you both for you input thus far. last night i tried the setting below, changes in bold, and did seem to be more stable...until i was playing a video and got a BSOD but i was able to run WCG longer.

Processor: Q9505
BIOS revision: 1802
Ai Overclock Tuner: Manual
Ratio CMOS Setting: 8.5
FSB Frequency: 412
CPU Clock Skew: Normal
NB Clock Skew: Normal
FSB Strap to North Bridge: 333
DRAM Frequency: DDR2-990MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 18 DRAM Clocks
RAS# to RAS# Delay: 3
Row Refresh Recycle Time: 55 DRAM Clocks
Write Recovery Time: 6
Read to Precharge Time: 3

Read to Write Delay (S/D): 8
Write to Read Delay (S): 3
Write to Read Delay (D): 5
Read to Read Delay (S): 4
Read to Read Delay (D): 6
Write to Write Delay (S): 4
Write to Write Delay (D): 6

Write to PRE Delay: 14
Read to PRE Delay: 5
PRE to PRE Delay: 1
All PRE to ACT Delay: 6
All PRE to REF Delay: 6
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: enabled
Ai Clock Twister: Moderate
Ai Transaction Booster: Manual

Common Performance Level: 12
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
...
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled
...

PCIE Frequency: 101

CPU Voltage: 1.30625
CPU PLL Voltage: 1.5265
FSB Termination Voltage: 1.3385
DRAM Voltage: 2.118
North Bridge Voltage: 1.3517
South Bridge 1.5 Voltage: 1.51325
South Bridge 1.1 Voltage: 1.11325

CPU GTL Reference (0): auto
CPU GTL Reference (1): auto
CPU GTL Reference (2): auto
CPU GTL Reference (3): auto
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto


CPU Configuration:

Ratio CMOS Setting: 8.5
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled

Load-Line Calibration: Disabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

After the BSOD, i went back in and tried -45mv GTL for 1 and 3, based on cstkl1's post http://www.xtremesystems.org/forums/showpost.php?p=3162579&postcount=331 but that didn't help at all, so i then tried +50 cause that makes more sense to me, my core 3 is the one that always failed Orthos so it would need more not less. I really don't understand the GTL settings. Based off the settings posted by Pink, we should be running the same GTLs since we have the same VTT voltage

[humeyboy]

Something far wrong there.

400FSB is not much an OC (stock on some older 65nm Extremes).

I have 475FSB on less voltages than you and @ 1:1 which is harder on the NB.

GTL's should not even be needed for that FSB, but sure its best to set them to the what you Calc them to be.

You certainly should not need touch Skews.

[64dragon]

i know, that's why i'm so confused. i'm thinking that when i get home from work tonight, i'm gona set it to 8x438 cause that gives me the same OC I had on the Xeon chip, maybe there is just something that the board/cpu/ram doesn't like about 412FSB. Voltage wise, the 9505 is the same as i had on the Xeon and VTT and vNB are lower, for the Xeon i needed 1.38 according to my notes

edit: forgot to mention but, when i was testing for max FSB, i had issues with .5 multipliers, when i set 6.5, it came up on post at 6 and in OS in Everest and CPUz. so, maybe my current instability is cause i'm using 8.5

[In-Fluence]

Have you triend bumping up the SB1.5 a bit, I read a few page back that it solved their problems, and if it happened whilst playing a video......worth a shot I reckon :)

Peace Mr.Pink - no offence taken - I was pretty tired at that point and probably over reacted a little, no offence taken or intended :)

p.s. looked at project dark matter, looking nice! Will have to tart my case up a bit when I've finalised a setup, but the possibility of a rampage extreme is also lurking. Here's a pic of mine with the custom cooling I mentioned earlier - nothing special:

http://i194.photobucket.com/albums/z182/in-fluence/DSCF0521.jpg

And btw, I'm womdering if everyone has these 'padded washers' either side of the NB? They seem a little thicker than the ones around the rest of the board and am wondering if Asus have attempted to resolve the poor contact issues they've had with the NB with this.

http://i194.photobucket.com/albums/z182/in-fluence/DSCF0507.jpg

[Mr.Pink]

Mr.Pink and In-Fluence, thank you both for you input thus far. last night i tried the setting below, changes in bold, and did seem to be more stable...until i was playing a video and got a BSOD but i was able to run WCG longer.



After the BSOD, i went back in and tried -45mv GTL for 1 and 3, based on cstkl1's post http://www.xtremesystems.org/forums/showpost.php?p=3162579&postcount=331 but that didn't help at all, so i then tried +50 cause that makes more sense to me, my core 3 is the one that always failed Orthos so it would need more not less. I really don't understand the GTL settings. Based off the settings posted by Pink, we should be running the same GTLs since we have the same VTT voltage

Just for the hell of it enabled Load Line Calibration and retest stability

[overclocking101]

I have this board and have hit ram clocks above 1200mhz on my ram therfore im guessing its the ram, it may be crap nb skew or not cant fix crap ram, but I may be wrong. Though with my crucial tracer REDS 2X2GB i got 1222mhz oc and they are rated stock at 800mhz

[Mr.Pink]

I have this board and have hit ram clocks above 1200mhz on my ram therfore im guessing its the ram, it may be crap nb skew or not cant fix crap ram, but I may be wrong. Though with my crucial tracer REDS 2X2GB i got 1222mhz oc and they are rated stock at 800mhz

My Dominators will do 1240MHz

http://i237.photobucket.com/albums/ff267/Grnfinger/1240ram_2.png

[Marcio]

http://img140.imageshack.us/img140/5180/xtremesystem.png

[overclocking101]

marcio what are you using to monitor your system temps/voltages?? is that everest or a seperate program?? also does anyone know what mofset waterblocks fit this board because my temp alarm on asus' pc probe keeps going off saying my power temp is at 50c or higher and I have the optional fan installed and another modded optional fan on the top mosfet heatsink so im just a bit confused

[Marcio]

everest, yes.

[eSp!s0]

Yes, its the Everest OSD.

[In-Fluence]

Just looking at your DDR1200 scores Marcio, and am wondering why my write and copy scores still seem to remain so slow (nice clock too btw). Copy is 1000MB/s off that score, and write is nearly 2000mb/s! I'm on PL7,Stronger, 5-5-5-15-45 at 450x8, and although I can probably tighten up the timings a little more and maybe lower the PL (still testing 7), I don't feel it would be enough to make that much difference, is this because of the difference between 45 & 65nm architecture?
p.s. I'm running the 400MHz strap - is this the cause? There's also 1200MHz available on the 266 strap, but couldn't post last time I tried.

Will request to place a work order for a Rampage extreme or a X48 equivalent today - any recommendations? Will hopefully get a Q9650 (or 9550) to match it too, and am looking at 3 sticks of DDR3 for which I'm open to suggestions too.

[Leeghoofd]

Are you using the exact same version of Everest In-Fluence ? There are sometimes "big" differences between different versions...

[Mr.Pink]

471FSB

http://i237.photobucket.com/albums/ff267/Grnfinger/Stable_471FSB-1.png

467FSB ( ram timing not tweaked)

http://i237.photobucket.com/albums/ff267/Grnfinger/4200_notweaks.png

500FSB 4.0GHz

http://i237.photobucket.com/albums/ff267/Grnfinger/500x8_1200ram.jpg

500FSB 4.5GHz

http://i237.photobucket.com/albums/ff267/Grnfinger/500FSB_MIIF.jpg

[CryptiK]

Just looking at your DDR1200 scores Marcio, and am wondering why my write and copy scores still seem to remain so slow (nice clock too btw). Copy is 1000MB/s off that score, and write is nearly 2000mb/s! I'm on PL7,Stronger, 5-5-5-15-45 at 450x8, and although I can probably tighten up the timings a little more and maybe lower the PL (still testing 7), I don't feel it would be enough to make that much difference, is this because of the difference between 45 & 65nm architecture?
p.s. I'm running the 400MHz strap - is this the cause? There's also 1200MHz available on the 266 strap, but couldn't post last time I tried.

Will request to place a work order for a Rampage extreme or a X48 equivalent today - any recommendations? Will hopefully get a Q9650 (or 9550) to match it too, and am looking at 3 sticks of DDR3 for which I'm open to suggestions too.

Although the board could probably handle PL6 at 450 FSB, it depends on the ram divider. The 266 strap will also require more voltage than the 333 so it might require a bump to post and a further increase to be stable.

I am running a Rampage Extreme currently, and I wouldn't advise you to pair it with a Q9650. From other guys results, you will probably struggle to get anything over ~475 FSB stable with the quad without using a lot of vNB. However if that's all you want out of it it will be fine. The REX is much better paired with a wolfdale if you want a flexible range of usable FSB. If you want the most out of a s775 quad, the UD3P/R really seem to be where its at.

I run Corsair 1866 7-8-7-20 (Two 2GB sticks of the 6GB triple channel kit) Elpida Hyper based ram on the REX. They are good, but if you want to push high MHz with tight timings and plenty of vdimm, a 2 x 1GB Micron D9 GTR/GTS kit is what you want.

[64dragon]

everything i tried with tuning the GTL's made me less stable,

tried:
-5,-45,-5,-45 and that lasted 19minutes in Orthos (core 4 failed)
10,-35,10,-35 lasted 9 minutes (core 4 failed)
i looked at the calculation again and saw that 10,-30,10,-30 was a closer match in voltage but that lasted 6 minutes (core 4 failed)
it seemed going this way was worse so i tried going the other way
-10,-55,-10,-55 lasted 5 minutes (core 1 failed), that threw a wrench in my theory so now i have no clue what to try
then i put the GTLs back on Auto and raised VTT and vNb 2 notches each and that lasted 1hr 30min (core 4 failed)

i really don't know what to do, this chip is pissing me off. this one might be going to my dad unless it clocks 10x better on a UD3 that i plan on getting for his pc (since the board is cheap and blue would clash with the red theme i have going)

edit: is there a software program that will display what the GTL values are in OS? if i were to set VTT and the GTLs back to auto and become stable, how could i find what BIOS was setting them to?

[humeyboy]

Have you tried Positive Values ?

[64dragon]

Have you tried Positive Values ?

not yet. using the equation, i got negative values so i stayed with that

[CryptiK]

AFAIK there's no app that can read the GTL's in windows. The manual doesn't exactly say what auto sets the Gtl's to and if this changes with an OC, so its anyones guess.

I would recommend you try positive values, the equation is to give people a starting point for further tweaking. Its also good for those that dont know how/dont have the inclination to tweak, but want to set the GTL's to 'something other than auto'.

You should just see what works best for your cpu and ignore what the equation says, as its not giving you the results you're after.

Try something like +50/+10/+50/+10 or higher even with Vtt set to ~1.20v for starters, perhaps increase Vtt to ~1.25v also and see what affect that has.

[humeyboy]

not yet. using the equation, i got negative values so i stayed with that

The Calcs are there as a starting point, you just 9 times out of ten make it Neg, I had to use Pos to get 475FSB.

My NB will not boot at Neg Values (well at overclocked speed) so its also Pos.

[In-Fluence]

Although the board could probably handle PL6 at 450 FSB, it depends on the ram divider. The 266 strap will also require more voltage than the 333 so it might require a bump to post and a further increase to be stable.

I am running a Rampage Extreme currently, and I wouldn't advise you to pair it with a Q9650. From other guys results, you will probably struggle to get anything over ~475 FSB stable with the quad without using a lot of vNB. However if that's all you want out of it it will be fine. The REX is much better paired with a wolfdale if you want a flexible range of usable FSB. If you want the most out of a s775 quad, the UD3P/R really seem to be where its at.

I run Corsair 1866 7-8-7-20 (Two 2GB sticks of the 6GB triple channel kit) Elpida Hyper based ram on the REX. They are good, but if you want to push high MHz with tight timings and plenty of vdimm, a 2 x 1GB Micron D9 GTR/GTS kit is what you want.

Yep, PL6 was a no-go, even going from PL7, stronger which was 10hours blend stable, to PL6, lighter wouldn't post. I'm happy to settle for that speed for now and can probably tweak a few more settings before calling it a day on that.
As for the REX, thanks for the advice. I've noticed the Gigabytes have been doing well in this area too so I will keep an eye on those. I quite liked the DFI LP UT NF4 SLI DR expert (or whatever it was) in the past, but it was very tricky to tweak, and so I am slightly dubious. DFI's warranty policy, sounds pretty strict from what I've heard too, but they've got the appealing option of being extremely tweakale.
Anyway, I may well still order a Gigabyte & 9650 and do a cheeky swap for my Q6600 to see how the 45nm quad performs on here, and maybe the mobo if it seems to do any better, but I do want to keep temps as low as possible for a near to silent system which makes me think the MIIF might be better. The whole DDR3 thing isn't quite so important for me right now and will probably try and stick with some combination of the aforementioned before I consider a more major upgrade in a year or two, but it would be interesting to get a feel for it.

Now payday's here, I've ordered a Corsair HX850, a Swiftech NB-Max block and 2 more feet of tygon tubing so I can sell off the MCW30 before it depreciates in value and hopefully improve the chipset/PWM cooling at little cost (and finally get rid of that awful bit of blue XSPC tubing and inefficient PSU). Good old tekheads have some great stock :)

One final thing (sorry to rant), now that I found a nice stable point for my system I tried lowering the vcore and re-tuning the GTLs, which have now made me think that they were'nt done properly (again)! I lowered the vcore to 1.425v (-.025v) whilst staying at 450x8 to make it unstable.

I tested with current GTLs (that I thought were the most stable) 50,10,50,10 and it reset in 2-3 minutes.
Then went 10,10,10,10 and 1024k paseed, and core 4 failed after 12 minutes of 8K (and gave a rounding error)
Going to 20,20,20,20 made 8K fail in about 2-3 minutes (followed by a reset)
So I tried -30,10,-30,10 as I had a hunch that perhaps GTLs 1 and 3 might be correctly tuned, and 0 and 2 were off and it passed 8K but almost immediately failed 10k and restarted
At -35,10,-35,10 10K ran for about 3 minutes longer before it restarted (but is the best so far)
Tested -25,20,-25,20 and it hung within 1 minute of 8K
Now testing -45,0,-45,0 (assuming auto is 0).
This time around, the results I'm getting from tuning the CPU GTLs feels 'right' whereas before results seemed a lot more random and no patterns were emerging (although I'm still rarely getting rounding errors, just reboots which still seems a bit off to me). Hopefully if I find one that stands out a lot more, I can start to push the CPU again, but I'm not budging until I'm certain of the best GTLs for me this time!

[humeyboy]

I don't get why you would need waste cash on WC for the Mobo.

They run far cooler (if not launch models, as Asus sorted heat issue with badyl mounted Heatpipes or you refit Heatpipes/TIM yourself) than say a Nvidia Chipset Mobo and hardly need any voltage to get to about 475FSB (1.26-1.28v NB) on a Quad and less for a Dual.

My NB and SB are about 43C and never go over 52C even in the summer with all fans at 50% and Priming or Intel Test Burning for 24 hours.

[In-Fluence]

I had the MCW30 block in the loop left over from my P5B setup and it seemed to help out on temps a reasonable amount once I managed to get a portion of the block on the mounting area (certainly did no harm).
I'm still having issues with the mosfet temps which seem to get worse when going beyond 1.45vcore and I normally get a reboot when they get to around 70°C. Using enzotech mosfet sinks without the heatpipe assembly resulted in even higher temps (all overheat protection bios options are disabled too), but both the NB & SB were noticeably cooler. I now have the heatpipe assembly back on it and it runs stable with the case open & house fan on it, so I'm hoping the NB-Max block will help out with this problem a little more by inducting that little bit more heat from the mosfet area, maybe not a great deal but it's of little cost and gives me peace of mind (unless I get a mosfet block).

I'm on 1.41v NB at the moment so it does run a little warm, but will try lowering it once I've got the CPU GTLs tuned.
As for those little critters, I've just tried -40,+10,-40,+10 and I got 1024k, 8k and 11 mins of 10k before a reboot, the best so far by about 8 minutes, but a gap of 50mv?!?!
Now trying -45,+10,-45,+10 to see if furthering this gap to 55mv offers any improvements - perhaps unorthodox going by the equation, but who knows?!
Forgot to mention I'm only running 1.32 vtt throughout all these tests so it would seem that 40/45mv should be the best gap.

[Haes]

not yet. using the equation, i got negative values so i stayed with that

no it's not negative values, cause you've to think in mathematical logic:

- * + = -

- * - = +

+ * + = +

so vtt*GTLref1-value=vtt*GTLref2

for example:

VTT=1.20 V

ideal GTLvalue 0.667% (or 0.67x)...

* = Quadcore needs a bit more max 0.7%
^ =Dualcore needs a bit less max. 0.635%

looking for idealvalue...

GTLref1= * for Quad, ^ for Dual
GTLref2 = ideal GTLvalue

1.2V*0.7x - value = 1.2V*0.667x

value = 1.2*0.7-1.2*0.667 = 1.2*(0.7-0.667) = +33 mV Quadcore
.................................................. ........... = -32mV Dualcore

[64dragon]

no it's not negative values, cause you've to think in mathematical logic:

- * + = -

- * - = +

+ * + = +

so vtt*GTLref1-value=vtt*GTLref2

for example:

VTT=1.20 V

ideal GTLvalue 0.667% (or 0.67x)...

* = Quadcore needs a bit more max 0.7%
^ =Dualcore needs a bit less max. 0.635%

looking for idealvalue...

GTLref1= * for Quad, ^ for Dual
GTLref2 = ideal GTLvalue

1.2V*0.7x - value = 1.2V*0.667x

value = 1.2*0.7-1.2*0.667 = 1.2*(0.7-0.667) = +33 mV Quadcore
.................................................. ........... = -32mV Dualcore

your equation is wrong, in bios the equation and using your #'s would be 1.2V*0.7x + value = 1.2V*0.667x
value = 1.2V*0.667x - 1.2V*0.7x = 1.2V*(0.667-0.7) = -.0396v = -40mv
i've been pluging in the equation to a ti-89 calculator and having it solve for my values...until i made my own excel sheet

either way, i tried:

50,10,50,10 and failed in 10 minutes on core 4
70,30,70,30 failed in 21 minutes on core 4 (my longes time up to this point)
80,40,80,40 failed in 5 minutes on core 4 (a bit worse than above run)
90,50,90,50 failed after 1hr 13min on core 4 (new record for longest run)
and just failed at 100,60,100,60 at 41 min on core 4 but with a fatal error instead of a rounding error like the rest. not sure what that means in relation to GTL tuning

so, do i go higher to 110,70,110,70 to see what happens or go back to 90,50,90,50 and then up my VTT and vNB until i'm stable? How high of a GTL value is safe? since i'm at .94v
my Xeon was 10 times easier to OC than this...but may have been able to lower voltages had i tried tuning the GTLs then.

[humeyboy]

See all that BS info in the Calc !, it is not even needed as it only confuses most peeps.


Example Quoted from elsewhere but far simpler than this sites :

FSB Termination: 1.35v
Formula:
Vtt x 0.667 + Item = Vtt x 0.635
1.35v x 0.667 + Item = 1.35v x 0.635
0.90045v + Item = 0.85725v
Item = 0.85725 – 0.90045
Item = - 0.0432v
Item= -43mv


Now round up to the nearest selectable value in the bios is -40mv and set the 0.667x Item value
to -40mv (negative), usually the (1 &3) core.
You have now made the 0.667x GTL voltages as close as possible to the 0.635x GTL voltages.

Bios setting Example:
CPU GTL REFERENCE (0) Auto
CPU GTL REFERENCE (1) -40mv
CPU GTL REFERENCE (2) Auto
CPU GTL REFERENCE (3) -40mv


Notice:
If the negative value is unstable then you might have to try the positive value and you also have to
tweaking it by trial and error until you find the stable value for your processor.



Now my way without Algebra crap.


Example :

FSB Term Voltage (VTT) = 1.2v

1) 1.2v x 0.667 = 0.8004

2) 1.2v x 0.635 = 0.762

3) 0.8004 - 0.762 = 0.0384 or 0.762 - 0.8004 = -0.0384

4) GTL's are +/- 0.38v so closest in our Bios is +0.40 or -0.35 or -0.40v (Neg are in 0.50v Steps, Pos are 0.10v in steps).

[Haes]

strange, that no matter which quad i had , all works great with "+" prime stable...

my Q6600 G0 30 10 30 10 30 3,6Ghz 1,343VC
my Q9550 E0 50 10 50 10 30 ~4,0Ghz 1,343VC
my Q9650 E0 20 10 20 10 30 4,2Ghz 1,328VC

could decrease 2-3 steps in VC everytime

and +30mv nb gtl (like already mentioned above) best stable NB for my in every fsb frequency

FSB VTT Temp

450 1.15 49°C
471 1.22 51°C
490 1.33 54°C max rock stable FSB
500 1.36 54°C bootable 15min stable
515 1.38 55°C bootable only 2 Cores, 3min stable

all with 2*4 gb RAM

nothing washed, all stock 1.02G A2 already about one year old


seriously, the only thing that bothers me, is the great Vdrop 0.08-0.09V.

[CryptiK]

@ Influence - There is a couple of DFI P45 boards that look like they perform ok with quads, but the best results I have seen are consistently on the UD3P or UD3R. DFI's are tweakable, but this does not always equate to better FSB or better performance. My DK X48-T2RSB Plus is a great board, especially for ram clocking (which is what I bought it for), but there are other X48 DDR2 boards that can achieve higher FSB etc. Depends what you want out of the board. There's plenty of good results in the UD3P thread and lots of guys using it, so getting advice & comparing settings is much easier than with the less popular DFI's.

@ Humeyboy - that equation is the same as the one I made in my GTL thread, just instead of 'item' I used 'Y'.

Here's one of mine for comparison (using Vtt of 1.14v):

Vtt x 0.667 + Y = Vtt x 0.635

1.14v x 0.667 + Y = 1.14v x 0.635
0.760v + Y = 0.724v
Y = 0.724v - 0.760v
Y = -0.036v
Y = -36 mv

@ 64dragon - you can always set the GTL's to the same mv offset too, like +50/+50/+50/+50, or decrease the gap, like +50/+40/+50/+40.

[humeyboy]

Yes but the extra non needed Algebra is BS and confuses many peeps even while at school (20+yrs ago) I remember most could not grasp it. :)

No need to make anything harder than it needs to be. ;)

Half can be totally skipped, sure it looks good on paper but its confusing to most.

[In-Fluence]

Well, I think for the price, Gigabyte UD3P range sounds appealing. I will research further.

The longest stable CPU GTLs I found were at -55,-5,-55,-5 which still doesn't seem right, but narrowing the gap gave me less stability (faill during 8K), while keping it at 50mv aways seemed to get into the 10k tests. Either way, it's still a big improvement over 50,10,50,10 which failed during 1024k, and I suspect I will be able to drop the vcore down by 2-3 notches.
One thing I did notice that seems to make me question if this is the right setting is that when I tested at +10,+10,+10,+10 and -20,-20,-20,-20, they gave me a rounding error, all others just rebooted (no BSOD either)

[CryptiK]

Yes but the extra non needed Algebra is BS and confuses many peeps even while at school (20+yrs ago) I remember most could not grasp it. :)

No need to make anything harder than it needs to be. ;)

Half can be totally skipped, sure it looks good on paper but its confusing to most.

I'm not sure how using the letter 'Y' as opposed to 'item' makes anything any harder to grasp, nor so I think it looks any better. You supplied an identical equation (maths wise) and described it as being 'way simpler than the one on this site' when the only difference is that it uses 'item' instead of 'Y'.

The whole equation working is there to show the process, of course once you get it you can skip whatever you want.


Well, I think for the price, Gigabyte UD3P range sounds appealing. I will research further.

The longest stable CPU GTLs I found were at -55,-5,-55,-5 which still doesn't seem right, but narrowing the gap gave me less stability (faill during 8K), while keping it at 50mv aways seemed to get into the 10k tests. Either way, it's still a big improvement over 50,10,50,10 which failed during 1024k, and I suspect I will be able to drop the vcore down by 2-3 notches.
One thing I did notice that seems to make me question if this is the right setting is that when I tested at +10,+10,+10,+10 and -20,-20,-20,-20, they gave me a rounding error, all others just rebooted (no BSOD either)

In my experience/opinion a reboot is less stable than a rounding error.

I'm not sure if a fatal error means its less stable than a rounding error. It sounds worse but who knows for sure.

Sounds like you need more vcore maybe more Vtt and just get it stable, then tweak the GTL's and try and lower vcore/vtt from that point.

[64dragon]

Hey guys, I think i'm acutally stable :shocked: i made the below settings last night and woke up this morning to a 14.5hr large FFT pass. clearly, ram was never my issue cause i'm finally running it OCed. so, my question now, what is strap and how does it affect things? I went from the 333 to 400 and became stable. Does strap affect "rated FSB"?

Processor: Q9505
BIOS revision: 1802
Ai Overclock Tuner: Manual
Ratio CMOS Setting: 8.5
FSB Frequency: 409
CPU Clock Skew: Normal
NB Clock Skew: Normal
FSB Strap to North Bridge: 400
DRAM Frequency: DDR2-1090MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 18 DRAM Clocks
RAS# to RAS# Delay: 3
Row Refresh Recycle Time: 55 DRAM Clocks
Write Recovery Time: 6
Read to Precharge Time: 3

Read to Write Delay (S/D): 8
Write to Read Delay (S): 3
Write to Read Delay (D): 5
Read to Read Delay (S): 4
Read to Read Delay (D): 6
Write to Write Delay (S): 4
Write to Write Delay (D): 6

Write to PRE Delay: 14
Read to PRE Delay: 5
PRE to PRE Delay: 1
All PRE to ACT Delay: 6
All PRE to REF Delay: 6
DRAM Static Read Control: Disabled
DRAM Read Training: Disabled
MEM. OC Charger: enabled
Ai Clock Twister: Moderate
Ai Transaction Booster: Manual

Common Performance Level: 12
Pull-in of CHA PH1: Disabled
Pull-in of CHA PH2: Disabled
...
Pull-in of CHB PH1: Disabled
Pull-in of CHB PH2: Disabled
...

PCIE Frequency: 101

CPU Voltage: 1.31
CPU PLL Voltage: 1.54
FSB Termination Voltage: 1.3385
DRAM Voltage: 2.118
North Bridge Voltage: 1.378
South Bridge 1.5 Voltage: 1.51325
South Bridge 1.1 Voltage: 1.11325

CPU GTL Reference (0): 90
CPU GTL Reference (1): 50
CPU GTL Reference (2): 90
CPU GTL Reference (3): 50
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto

CPU Configuration:

Ratio CMOS Setting: 8.5
C1E Support: Disabled
Max CPUID Value Limit: Disabled
Intel Virtualization Tech: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled

Load-Line Calibration: Disabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

[humeyboy]

The FSB Strap gives you Memory Ratios to whatever FSB you set but it also changes Timings.

So 400 is slacker than 333 so your Memory Bandwidth will be lower but its easier on the NB.

Install latest Bios FFS.

Then set to 400FSB, 333 Strap and Memory either Auto or 400 and Moderate + PL 10 for 1:1.

It should work with very little voltages.

[Mr.Pink]

Hey guys, I think i'm acutally stable :shocked: i made the below settings last night and woke up this morning to a 14.5hr large FFT pass. clearly, ram was never my issue cause i'm finally running it OCed. so, my question now, what is strap and how does it affect things? I went from the 333 to 400 and became stable. Does strap affect "rated FSB"?

Congrats on the stable settings...
There are much better ram settings than what your running, PL12 is excessive for that clock/strap.

[Haes]

@mr pink:

can you give your bios settings for 4.2ghz 9650 pls. I need to compare something... thx

[64dragon]

my "stability" deceived me, woke up this morning to a BSOD and 2 work units in WCG had computation errors, could be the ram since it is now OCed

[Haes]

My PC is empty at the moment since I'm waiting for my Q9650 to arrive so I took the opportunity to remove the motherboard heatsink and replace the crap that ASUS put on with my Zalman thermal paste. I was getting NB/SB temps of 45-48C with everything at stock so hopefully that'll improve. What sort of chipset temps are you guys getting, particularly those of you who have replaced the TIM with something better?

Lenny, can you give your bios settings too pls.

[Mr.Pink]

@mr pink:

can you give your bios settings for 4.2ghz 9650 pls. I need to compare something... thx

Certainly :D

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 467
CPU Clock Skew : Delay 100ps
North Bridge Clock Skew : Delay 100ps
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-1121
Dram Clock Skew CH1 A1 : Normal
DRAM Clock Skew CH1 A2 : Normal
Dram Clock Skew CH1 B1 : Normal
Dram Clock Skew CH1 B2 : Normal


DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time : 55
Write Recovery Time : 6
Read to Precharge Time : 3

Read to Write Delay (S/D) : 8
Write to Read Delay (S) : 3
Write to Read Delay (D) : 5
Read to Read Delay (S) : 4
Read to Read Delay (D) : 6
Write to Write Delay (S) : 4
Write to Write Delay (D) : 6

Write to PRE Delay : 14
Read to PRE Delay : 5
PRE to PRE Delay : 1
ALL PRE to ACT Delay : 5
ALL PRE to REF Delay : 5

DRAM Static Read Control: Enabled
Dram Read Training : Disabled
MEM OC Charger : Enabled
Ai Clock Twister : Stronger
Transaction Booster : Manual

Common Performance Level [8]

Pull-In of CHA PH1 Disabled
Pull-In of CHA PH2 Disabled
Pull-In of CHA PH3 Disabled
Pull-In of CHA PH4 Disabled
Pull-In of CHA PH5 Disabled
Pull-In of CHB PH1 Disabled
Pull-In of CHB PH2 Disabled
Pull-In of CHB PH3 Disabled
Pull-In of CHB PH4 Disabled
Pull-In of CHB PH5 Disabled

PCIE Frequency : 103

CPU Voltage : 1.325
CPU PLL Voltage : 1.59275
FSBT : 1.37825
DRAM Voltage : 1.9
North Bridge Voltage : 1.4108
South Bridge Voltage 1.5 : 1.55300
South Bridge Voltage 1.1 : 1.20600

CPU GTL Reference 0 : Auto
CPU GTL Reference 1 : -45mv
CPU GTL Reference 2 : Auto
CPU GTL Reference 3 : -45mv
North Bridge GTL Reference : AUTO
DDR2 Channel A REF Voltage : AUTO
DDR2 Channel B REF Voltage : AUTO
North Bridge DDR Reference : AUTO

Load Line Calabration : Enabled
CPU Sread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

[Haes]

Certainly :D

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 467
CPU Clock Skew : Delay 100ps
North Bridge Clock Skew : Delay 100ps
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-1121
Dram Clock Skew CH1 A1 : Normal
DRAM Clock Skew CH1 A2 : Normal
Dram Clock Skew CH1 B1 : Normal
Dram Clock Skew CH1 B2 : Normal


DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time : 55
Write Recovery Time : 6
Read to Precharge Time : 3

Read to Write Delay (S/D) : 8
Write to Read Delay (S) : 3
Write to Read Delay (D) : 5
Read to Read Delay (S) : 4
Read to Read Delay (D) : 6
Write to Write Delay (S) : 4
Write to Write Delay (D) : 6

Write to PRE Delay : 14
Read to PRE Delay : 5
PRE to PRE Delay : 1
ALL PRE to ACT Delay : 5
ALL PRE to REF Delay : 5

DRAM Static Read Control: Enabled
Dram Read Training : Disabled
MEM OC Charger : Enabled
Ai Clock Twister : Stronger
Transaction Booster : Manual

Common Performance Level [8]

Pull-In of CHA PH1 Disabled
Pull-In of CHA PH2 Disabled
Pull-In of CHA PH3 Disabled
Pull-In of CHA PH4 Disabled
Pull-In of CHA PH5 Disabled
Pull-In of CHB PH1 Disabled
Pull-In of CHB PH2 Disabled
Pull-In of CHB PH3 Disabled
Pull-In of CHB PH4 Disabled
Pull-In of CHB PH5 Disabled

PCIE Frequency : 103

CPU Voltage : 1.325
CPU PLL Voltage : 1.59275
FSBT : 1.37825
DRAM Voltage : 1.9
North Bridge Voltage : 1.4108
South Bridge Voltage 1.5 : 1.55300
South Bridge Voltage 1.1 : 1.20600

CPU GTL Reference 0 : Auto
CPU GTL Reference 1 : -45mv
CPU GTL Reference 2 : Auto
CPU GTL Reference 3 : -45mv
North Bridge GTL Reference : AUTO
DDR2 Channel A REF Voltage : AUTO
DDR2 Channel B REF Voltage : AUTO
North Bridge DDR Reference : AUTO

Load Line Calabration : Enabled
CPU Sread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

thx m8 appreciate this

[In-Fluence]

In my experience/opinion a reboot is less stable than a rounding error.

I'm not sure if a fatal error means its less stable than a rounding error. It sounds worse but who knows for sure.

Sounds like you need more vcore maybe more Vtt and just get it stable, then tweak the GTL's and try and lower vcore/vtt from that point.

This is where I experience trouble finding the GTLs though. If I up Vcore a notch or two, I will get more stability (if not, total stability), but it will still reboot on the majority of occasions. If I give more or less VTT than 1.32 then the system becomes more unstable. On another note, I got back after a weekend away, started the rig up, and with the most stable GTLs I found so far, it was failing withing 5-10 minues of prime (during 1024k test) which it passed all the time 2 days ago and was stable for almost an hour. Something very fishy here imo.