ASUS Maximus II Formula - new P45 king?

[Grnfinger]

I'm very tempted to get this motherboard... I just got a P5Q-E and the lack of sensors on the motherboard is sorely disappointing. The only voltage value I can view in windows is vcore don't get important voltages like DIMM, NB, SB, VTT

I know that the MF2 is regarded by many as "too expensive" for its raw performance, but dayum it's the best looking mobo I've ever seen in my life

While others struggle with the board, its actually not a bad board, OC'ing quads is a little harder than dualies, it requires alot of tweaking its very fussy on settings, too much or too little and it will spit in your face while laughing at you.
If the bios ever matures the board will truely shine, comparing an X38 Maximus and a X48 Rampage the board hangs in ther very well....Lotta UD3P fanbois out there, yeah it oc's quads very well but it comes with a price

I've tested 4 chips ( all quads) and have 2 Maximus II Formula boards all chips and both boards will do 500FSB

DON'T.

Buy the Gigabyte, this Mobo is crap for Quad's.

I have just swapped my E8600 for a Q9650 and cannot even get 450FSB 100% stable.

CPU does 3GHZ with 1.03v which proves the VID is a lot of crap as mines is 1.275v (high) this is Primed Stable/Vantage etc).

For now I'm at 9x400 with all lowest Mobo voltages, but I cannot get 450 stable so far and 500FSB will not even boot into Windows no matter how much voltage etc.

Mobo is a Rev A2, doubt the Rev A3 NB helps that much, Asus sure have made a boob here.

We have same board same chip your ram is different but the timings are the same, start with 4GBs and try this...it should help point you in the right direction

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

[LennyRhys]

UD3P is a great board but so is the MF2... the very first time I saw the MF2 I thought "I will own this motherboard one day, even if it kills me"

It's just one of these things *that you never tell people about*

There is only one problem...it seems that every retailer in the UK has 0 boards in stock

[In-Fluence]

Specialtech have them, it's where I got mine, and it came unopened, with all the protective stickers in place, basically, definitely unused
http://www.specialtech.co.uk/spshop/...cat=903&page=1

Pretty good price too

Anybody care to help me re: my previous post re: Simp's guide?

[Grnfinger]

Back again! Managed to get the board fitted & chugging along nicely after some careful removal of the giant heatsink layout, installation of the NB block, and some copper on the remaining important parts. Temps are so much lower on this than my P5B it's unreal, and it's nice to get similar stability on lower volts too! It now makes me think I've got a better Q6600 than I originally thought, but this board is a bit of a beast, and it's bringing me back to the days of my DFI NF4-SLI-DR Expert (or whatever it was which I spent a looong time tweaking!

Anyway, I digress - after going through Simp's superbly written guide, I have at least started to mke some headway veering towards the 3.8GHz mark using 418x9 with memory on the 400 divider (running 1179MHz MT stable), although I do have a couple of questions about the methods for testing..

• When you say 'prime unstable' I'm going by the point at which the computer just restarts. I'm almost assuming this is wrong, as I clearly wouldn't be able to make note of the time a test failed, (yes, I've been sitting there waiting for reboots). I got as far as the 2nd point in the guide before I thought I should ask before I possibly waste what little time I have! Results to seem reasonably consistent, but I fear could be a result of my power temp going too high is ~60°C too much? And could this represent the
  mosfet temps? Should I be trying to aim for rounding errors?
  
• One thing that raised my attention was the fact that the standard blend test goes from a high mem/nb test (1024k) straight to a high CPU stress (8k). Would it not be better to try and run more relevant tests when testing the results of certain settings? i.e. when adjusting CPU GTLs, running small fft tests to test CPU stability, and when adjusting NB GTLs/RAM skews, go for higher ones to test NB & RAM? Perhaps blends would be better suited to the CPU/NB delays?
  
• Using the voltages Simp provides at the beginning of the guide, and my own ideal 24-hour vcore & vdimm values, I get reboots as mentioned before. Would starting from a lower clock and slightly undervolting to try & provoke rounding errors be better? The guide says I can adjust the FSB later on, but I'm wondering if any dramatic change would just mean I would have to re-tune everything again?
  

Sorry for the rant guys, I just wanna get this right and determine if I'm missing something silly, or if a certain component might be letting me down!

Stability
Always start with prime small FFT's this will test teh cpu only....once you get it stable then run blend it will stress the MCH and cpu...rounding errors are a sign of a unstable rig...reboots of course is a very unstable rig.

Temps
60c is ok for your Q6600 dont let it get over 70c for long periods of time

As for Simps guide I read it and tested it out on my second MIIF board, I did not have any higher clocks from my own method...not overly familar with it so perhaps someone else could comment about that.

UD3P is a great board but so is the MF2... the very first time I saw the MF2 I thought "I will own this motherboard one day, even if it kills me"

It's just one of these things *that you never tell people about*

There is only one problem...it seems that every retailer in the UK has 0 boards in stock

I know a Canadian who has a gently used MIIF for sale

[In-Fluence]

Thanks Grnfinger - just what I hoped to hear

now all I need to do is find an unstable setting that gives only rounding errors - not having a lot of luck so far - I can't help but get reboots after 10 or so minutes of small or large prime tests Any hints?

p.s. CPU temps are welll within limits - it was just the power / pwn temp I'm a little concerned about (although re-seating the mosfet sinks seem to have helped a fair bit now).

Thanks again, I'll be sure to post settings once I'm all tuned up

[CryptiK]

@ In-Fluence - sounds like you have picked an arbitrary speed and are trying to stabilize things there. The best way is to work your way up, adjusting one thing at a time as you reach a point of instability.

[In-Fluence]

You're quite right Cryptik, I've been trying to run before I can walk!

I was going by my o/c on the P5B-dlx and was just hoping I would be able to get near to that without having to do too much else, although it seems this beast needs to be well and truly tamed, particularly with quads.
I did manage to get around 5hrs prime blend stable after going down to 400FSB a few days back (i stopped the test), but notching up the FSB 1MHz at a time after this only resulted in the usual 10-minute reboots. It's the timing that gets me, reboots seem to happen within 3-15 mins of each test..makes me think something might be overheating

I'm also getting a few hung POSTS, along with 'DET_DRAM' and 'POWER_OFF' messages on the LCD occasionally too. I hear this might be related to a bad bios flash, so I'm going to attempt a reflash in proper DOS mode tonight.

Failing all that, I'll test at stock speeds again, and if that fails, it might be an RMA...I'm beginning to grow a tad suspicious of this board, even though I have some sentiment towards Asus products!

Thanks for the help already though, hopefully I'll make some good headway tonight... Wish me luck!

[CryptiK]

You're quite right Cryptik, I've been trying to run before I can walk!

I was going by my o/c on the P5B-dlx and was just hoping I would be able to get near to that without having to do too much else, although it seems this beast needs to be well and truly tamed, particularly with quads.
I did manage to get around 5hrs prime blend stable after going down to 400FSB a few days back (i stopped the test), but notching up the FSB 1MHz at a time after this only resulted in the usual 10-minute reboots. It's the timing that gets me, reboots seem to happen within 3-15 mins of each test..makes me think something might be overheating

I'm also getting a few hung POSTS, along with 'DET_DRAM' and 'POWER_OFF' messages on the LCD occasionally too. I hear this might be related to a bad bios flash, so I'm going to attempt a reflash in proper DOS mode tonight.

Failing all that, I'll test at stock speeds again, and if that fails, it might be an RMA...I'm beginning to grow a tad suspicious of this board, even though I have some sentiment towards Asus products!

Thanks for the help already though, hopefully I'll make some good headway tonight... Wish me luck!

Use a template and fill out your complete BIOS settings, we can go from there. You should easily be able to get 450 FSB stable with the Quad. It's 450 - 500 FSB where it seems to be hard to stabilize.

At what speed/vcore/Vtt were you running that CPU stable at on the P5B-Dlx?

Check your temps when it's crashing - is anything overheating? I have had PWM temps in the 50's and NB temps in the 60's without it causing errors, so they can take a bit of heat, but lower temps are always better than higher ones. Some boards may be more heat sensitive than others though.

Personally I think you may benefit from leaving your chip at 6x multi, and clocking up the FSB gradually so you can focus mainly on the board and ram. Then bring the cpu back into play with a higher multi once you have a given FSB/ram speed stabilized. Leave everything at default voltage (but set them manually of course) except vNB for which you can set ~1.30v. Since you know you can get 400 FSB stable start there. Work your way up and when it becomes unstable, tweak things one at a time until you find what the limiting parameter is. Then move up again etc.

Your PSU could also be causing instability over a certain power draw, I have never heard of that brand you use but if all else fails perhaps it would be worth borrowing one off a mate to test the theory.

Interesting your board is an A3 rev. chipset, first M2F I have heard of with one.

[In-Fluence]

Will do, was busy rebuilding it last night just so I could re-seat the mosfet sinks and rule out them overheating, I have improved on the NB temps too which don't hit 50 now. (I thought it'd be lower with a swiftech block on it) :/

Anyway, here's a snapshot of it testing at 400x9 with temps, volts and chipset info:
http://i194.photobucket.com/albums/z...ce/Testing.jpg

I'll get on with testing FSB at lunch or after work. PSU does have a few years of use although it was pretty well rated back then, one of the first 'unofficial' SLI PSUs which was running everything I have now plus one more 8600GT - and it weighs quite a bit which I know is often a good sign

[LennyRhys]

How safe is it to remove chipset cooling on the M2F? I did it with my 680i board and replaced the thermal gunk with my Zalman paste and it made a world of difference, so I plan on doing the same thing to my Maximus...as long as removing the heatsinks won't damage the nekid chipset.

[WhiteFireDragon]

nope, won't damage it. it's actually advised that you redo the TIM. once you take off the heatsink assmebly (it's only one big piece btw!), you'll notice how horrible they apply the paste on some boards.

[LennyRhys]

Ah ok cool - provided my board arrives tomorrow I'll get busy with a screwdriver and my cherished Zalman TIM, the best I've ever used. My evga 680i TIM was like a ball of play-doh, can't imagine I'll ever see worse than that lol.

[In-Fluence]

Think I sussed out what was going on. The VTT & Northbridge voltage difference bug seems to be coming into play bigtime here, I'm yet to find out the details, but I think the acceptable gap for me is something like 0.05 - 0.07v. Working within this limit isn't giving me any reboots now

I've now managed to save an OC profile that is 2hr small FFT stable at 450, so I'm just nudging up the FSB until I find the limit.
Currently testing at 454x6 now though, and seems to be going ok after 30 mins, NB@47°C on 1.3v, and VTT on 1.25v, power temp looming around 58°C... still having trouble finding a consistent rounding error point though.

Man this is a quirky board with this quad, I set the FSB to 452x6 and it kept booting at 448x6!

Once I do get there I'll post settings up - eveything's now starting to behave as I thought it should

[In-Fluence]

Decided to put the heatpipe assembly back on with my swiftech block, and temps seem to be going down a lot on the SB and a little on the PLL although the NB is a few degrees higher without the direct waterblock contact, but not really suffering too much. 454x6 seemed to test well for 3 hours or so, but it crapped out on 457..

I got a bit sidetracked too, and went for a couple of suicide runs at 4GHz+:



And the highest I dare go, but crashed during 32M:


I know these speeds aren't anywhere near stable but it's nice to see that my CPU could do it

Anyway, I'm getting back to finding my highest stable o/c, but still not having a great deal of luck avoiding the inconsistent stability. It appears to be small fft stable for a few hours, then after rebooting with the same settings and testing again, it's not! Either way, I'm convinced the board is actually ok, and it's just those voltages that are awry. I've got bios 1901 & 1204 onboard, but am contemplating 1307 - think it's worth it?

Here are my maximum stable FSB settings, any hints on going higher will be gladly taken on board

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 454
CPU Clock Skew : Auto
North Bridge Clock Skew : Auto
FSB Strap to North Bridge : 333
DRAM Frequency: DDR2-1091
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 : 18
RAS# to RAS# Delay : 3
Row Refresh Cycle Time : 55
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) : 7

Write to PRE Delay : 13
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 : Enabled
Ai Clock Twister : Moderate
Transaction Booster : Manual

Common Performance Level [10]

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 : 1.4
CPU PLL Voltage : 1.55300
FSBT : 1.25900
DRAM Voltage : 1.215
North Bridge Voltage : 1.29875
South Bridge Voltage 1.5 : 1.50000
South Bridge Voltage 1.1 : 1.10000

CPU GTL Reference 0 : Auto
CPU GTL Reference 1 : Auto
CPU GTL Reference 2 : Auto
CPU GTL Reference 3 : Auto
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

[Grnfinger]

1307 is a decent bios bios 1901 is very good, I found that 1204 was a real dog for my setup and quickly got rid of it.
If your passing small ffts what prime blend test are you failing 1024 --896 --- 768 ????
Try playing with clock skews they help a lot for higher FSB clocks

[In-Fluence]

Wow, that was actually a pretty golden hint in my case, thanks dude! I've tested the CPU/NB clock skews from auto and have moved up to 100/300ps delays which seem to be the most stable by far whilst running 467x6 I'm 30mins small fft stable so far (which I couldn't even boot at before). Can these delays remain the same at any FSB speed once tuned? Going up to around 490 fsb before always resulted in 2 cores getting disabled, could these skews be responsible for this? I guess I'll have to try & see for myself!

I haven't got around to doing the blend tests just yet, as I'm still trying to find my max CPU FSB which has now just got a big boost. Once I get there, I'll start pushing the memory & NB and doing the more relevant Blend tests.

Sorry I have so many questions, I've been a keen O/cer for a few years now and although I'm patient, I'm eager to get this thing running optimally without wasting too much time.

Thanks for the tips so far

[Grnfinger]

properly tuned GTL's and Skews are the secret to this board, I leave my skews set at teh same values for 467FSB and 500FSB, testing each is a long process but well worth the effort in the end... looking forward to your results

[In-Fluence]

Will be certain to post results, thanks for your input. I think the order in which I was trying to do things (i.e. as with simp's guide) might have been hindering me whilst trying to get over 460. The current CPU/NB skews have clearly allowed me to up the FSB significantly.

Still having trouble getting over 470 though - it seems like a die is getting disabled. It boots & primes ok at 500 on the 2 cores for a few mins with memory at 1200. I'm suspecting the GTL refs, but I've tried a few now to no avail, along with high & low voltages on vcore, vtt, nb, pll in several combinations (and a couple of 'known' NB GTLs that helped me in the past). The moment I go back to 470, it picks up all 4 cores again

I think I'll find the FSB point at which it boots with 2 cores and see if helps to make any changes from there, but once again, I'm open to suggestions

[eSp!s0]

Just started to test for minimal voltages with my E8400 @ 533x7,5, ram 1:1 (333Strap).

Alreay finished VNB and need 1,285V for 533Fsb 1:1 (1,179VNB for 500Fsb 1:1, 333Strap too)

[Grnfinger]

nice I need at least 1.41vNB for stability with my quad running 12:10 1121MHz with
DRAM Static Read Control: Enabled
Dram Read Training : Disabled
MEM OC Charger : Enabled
Ai Clock Twister : Stronger
Transaction Booster : Manual

Common Performance Level [8]

[LennyRhys]

Got my MF2 today, I bought it as new but it came in a box with an RMA sticker (typical ebay) but it seems to work fine. I won't be overclocking with it until my Q9650 arrives.

[eSp!s0]

nice I need at least 1.41vNB for stability with my quad running 12:10 1121MHz with
DRAM Static Read Control: Enabled
Dram Read Training : Disabled
MEM OC Charger : Enabled
Ai Clock Twister : Stronger
Transaction Booster : Manual

Common Performance Level [8]

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

Common Performance Level [10]

[Grnfinger]

Nice bandwidth scores

[In-Fluence]

Ok, after countless tweaks, reboots, CMOS clears and so on, I've finally managed to get to grips with the board. I'm now at the point where I've properly tuned the GTLs, (+40,0,+40,0) and clocked to the highest speed I can (and tuned the memory a little), and was pleasantly surprised to wake up and find it still testing



As you can see, the temps are at the absolute limit (it shuts off when CPU reaches 75°C), but considering temps in my flat are ranging from 28-35°C at the moment, I don't think I'm doing too bad! Hopefully after tuning the NB GTL and clock skews I'll either be able to drop vcore or raise FSB a fair bit.
I'm guessing that an IBT run will probably raise my temps some 10-15°C which isn't really an option in this heat, but I'll get around to it when it decides to cool down, I really don't wanna toast the CPU!

So what do you think of the results so far (particularly for memory)?

After I've done a complete tuneup of all settings at this multi / divider, I'll probably move onto 400x9 and do the same, and compare results

[Grnfinger]

bandwidth looks decent altho your copy score looks low