ASUS Maximus II Formula - new P45 king?

[CryptiK]

I can run 3900 10 hours stable with reasonable volts.
333 strap and 1103MHz ram PL8, LLC enabled

My GTL's are set to a + value any change and it fails prime and linpack
The values dont match the formula I know but they work so I dont argue with them

CPU Voltage : 1.3875
CPU PLL Voltage : 1.53975
FSBT : 1.33850
DRAM Voltage : 2.095
North Bridge Voltage : 1.39150
South Bridge Voltage 1.5 : 1.5
South Bridge Voltage 1.1 : 1.1

CPU GTL Reference 0 : +30mv
CPU GTL Reference 1 : +20mv try +30/+40/+50mv
CPU GTL Reference 2 : +20mv try +30/+40/+50mv
CPU GTL Reference 3 : +30mv
North Bridge GTL Reference : AUTO
DDR2 Channel A REF Voltage : AUTO
DDR2 Channel B REF Voltage : AUTO
North Bridge DDR Reference : AUTO

If core 2 always fails, try increasing the GTL Ref for that core. I believe it is 'core 1' (as 4 cores are denoted 0,1,2,3) but also try increasing it on 'core 2' it *should* make a difference.

[Copenhagen69]

whats highest stable FSB anyone is getting on this board?

i got 480fsb but with 2.25V (2.1 stock) i cant seem to break the 500FSB mark ... has anyone? is it common?

using 1066 dominators

[The Nemesis]

whats highest stable FSB anyone is getting on this board?

i got 480fsb but with 2.25V (2.1 stock) i cant seem to break the 500FSB mark ... has anyone? is it common?

using 1066 dominators

I topped out @ 495FSB boot from windows. Can use setFSB or aisuite to get to 500FSB but its not stable. Few have gotten this board or any other for that matter to reach 500FSB on a 45nm quad. I finally quit this weekend and settled on 475 X 8 1.35vcore.

[Seewolf]

500 Mhz are possible with my Q9550



I need 1,5V NB and I can boot @ 500 Mhz and Bench 3D Mark and all other Benchmarks.

[CryptiK]

whats highest stable FSB anyone is getting on this board?

i got 480fsb but with 2.25V (2.1 stock) i cant seem to break the 500FSB mark ... has anyone? is it common?

using 1066 dominators

You must mean with quad's, right? And what is stable to you? Orthos, SPI32M etc.

[GaBBa-Gandalf]

Hello everybody,

i heard that the voltages from the board are extremly high - also if u use the lowest voltage settings. This should happen if u use a fsb over 450 MHz.

I would like to check this... Do anybody know where are the measure points for different voltages of this board?

[Copenhagen69]

You must mean with quad's, right? And what is stable to you? Orthos, SPI32M etc.

nope im running e8400
well stable coule be long enough to run some benchies or 24/7

[Starage]

whats highest stable FSB anyone is getting on this board?

i got 480fsb but with 2.25V (2.1 stock) i cant seem to break the 500FSB mark ... has anyone? is it common?

using 1066 dominators

Hey what voltages are you talking about? What is 2.25v for.
If you are running an E8400, I have hit 560 FSB, but ddr2-1066 is only good for 533. I have tried 570 and 588 but at that bus speed is not stable.
I am using a 7x multiplier for a 3920mhz core speed. My E8400 is not stable over 4000mhz. My DDR2-1200 is running at 1:1 or (560mhz) 1120.

[vaderat-at]

550fsb here. 3 hours and counting. I'll let it run for 12 to make sure. I had to up the vcore slightly to achieve this.

[lowdog]

Correct except for your conversion from volts to millivolts. It's not -12.5mv, it's -125mv

There's 1000 mv in 1 volt

0.001v = 1 mv
0.010v = 10 mv
0.100v = 100 mv



**There's something I want to add concerning this, I'm relatively sure the Reference voltage should scale up with increasing vdimm, however I'm not sure what the 'safe max' is, so that's why I dropped it back a little, but I'd like some input from anyone who knows something about it. Where's freecableguy when you need him - I may ask about it on the tech repository.

Can you please let us know if and when you find out what and how to set these GTL's - NB GTL, DDR2 chA and DDR2 chB, NB DDR Ref.

At present with this board I have my Q9650 running at 3.6GHz, 9x400fsb with:

Cpu v - 1.2v with LL disabled = 1.16v idle and 1.12v load (Linpack and Prime95 stable)
CPU PLL - 1.51v
FSB Term V - 1.118v
Dram V - 1.82v (8gb of Corsair Dominator 8500C5DF)
NB V - 1.118v
SB 1.5 - 1.51v
SB 1.1 - 1.10v

CPU GTL (0) auto
CPU GTL (1) -40
CPU GTL (2) auto
CPU GTL (3) -40

All other GTL's auto

LL - disabled.

Going to push for 4GHz but I want to know what all the other GTL's do and how to set them first before I push up volts etc..

On stock speed of 333fsb all voltage setting when set to auto seem to give minimum volts across the board, when the fsb is bumped to 400 all auto volts SKYROCKET to dangerous levels, so people had better be carefull when running 45nm to make sure they are manually setting ALL volts when they start overclocking and trying to keep whatever and if not all volts as minimal/low as possible if they can keep stability like that.

[vab206]

im getting 4ghz easily with 450fsb X 9, ram set to 1199 @ 5-5-5-12. was booting at 500fsb but had to drop my ram down considerably to do so, went with 450 in the end.

[CryptiK]

Can you please let us know if and when you find out what and how to set these GTL's - NB GTL, DDR2 chA and DDR2 chB, NB DDR Ref.

At present with this board I have my Q9650 running at 3.6GHz, 9x400fsb with:

Cpu v - 1.2v with LL disabled = 1.16v idle and 1.12v load (Linpack and Prime95 stable)
CPU PLL - 1.51v
FSB Term V - 1.118v
Dram V - 1.82v (8gb of Corsair Dominator 8500C5DF)
NB V - 1.118v
SB 1.5 - 1.51v
SB 1.1 - 1.10v

CPU GTL (0) auto
CPU GTL (1) -40
CPU GTL (2) auto
CPU GTL (3) -40

All other GTL's auto

LL - disabled.

Going to push for 4GHz but I want to know what all the other GTL's do and how to set them first before I push up volts etc..

On stock speed of 333fsb all voltage setting when set to auto seem to give minimum volts across the board, when the fsb is bumped to 400 all auto volts SKYROCKET to dangerous levels, so people had better be carefull when running 45nm to make sure they are manually setting ALL volts when they start overclocking and trying to keep whatever and if not all volts as minimal/low as possible if they can keep stability like that.

I know, and have briefly explained how, to set the GTL Ref's on this board. Depending on how each core of your CPU behaves, you need to tweak them yourself accordingly, after setting the standard values, which as I said earlier, are 0.635x Vtt for 45nm CPU's.

This means you need to adjust GTL Ref 1 & 3 (0.667x) to a negative mv value (which you work out by using the equation or second, easier method I gave based on your Vtt) as you have done, and leave the GTL Ref 0 & 2 (0.635x) alone so all the GTL Ref values are as close to the same as you can get them.

From that point you need to tweak for your individual CPU, GTL's are a not a 'set this number and you're done' setting. For example your core 0 may be weak, so it may need more voltage than the others etc.

For the NB GTL Ref, I have had no reason to change it from stock yet, it may help get the FSB to 600+ levels but I haven't had time to play with it yet.

The max FSB on my board I can get into windows with is 550 (1.30v vNB) so something isn't right there, as my CPU will do 600FSB on 1.344v, I could do 600FSB at 1.26v vNB on my P5Q-Deluxe easily.

I think the M2F bios needs to be much better optimized for high FSB and also high ram speeds, I could do much better on my P5Q-Deluxe in both respects. Disappointing considering the price of the P5Q-Deluxe compared to the M2F. Also, the M2F is way too sensitive to the vFSB:vNB voltage ratio, requiring vNB to be no more than 0.113v higher than vFSB or the board will not POST on reboot - again the P5Q-Deluxe was better and allowed you to run ~1.34v vNB and 1.11v vFSB without issues. I'm sincerely hoping the next M2F BIOS revision will address these issues.

[lowdog]

I know, and have briefly explained how, to set the GTL Ref's on this board. Depending on how each core of your CPU behaves, you need to tweak them yourself accordingly, after setting the standard values, which as I said earlier, are 0.635x Vtt for 45nm CPU's.

This means you need to adjust GTL Ref 1 & 3 (0.667x) to a negative mv value (which you work out by using the equation or second, easier method I gave based on your Vtt) and leave the GTL Ref 0 & 2 (0.635x) alone so all the GTL Ref values are as close to the same as you can get them.

From that point you need to tweak for your individual CPU, GTL's are a not a 'set this number and you're done' setting.

Yeah I understand that in regard to the individual CPU GTL's and have set them accordingly (spelling ) and can comprehend the formular for working out the correct setting.

My question was in reference to the NB GTL , NB DDR Reference V and DDR2 ChA/B Reference V. What they do? how important/crucial are they? are there any adverse affects leaving them on auto? etc. Would be nice to get some difinitive answer regarding formular on setting these correctly and their function in so called "laymens terms"

[CryptiK]

Well I have PM'd FCG, hopefully he can shed some light on it, as I'm unsure what to recommend regarding these voltages. I'll let you know if he gets back to me, and in the meantime I'll do some research tonight when I get home and try to figure out what they are/do.

[lowdog]

Well I have PM'd FCG, hopefully he can shed some light on it, as I'm unsure what to recommend regarding these voltages. I'll let you know if he gets back to me, and in the meantime I'll do some research tonight when I get home and try to figure out what they are/do.

Coooooool, and thanks. Will hopefully help us all out if we get some light shedded on these settings.

[Grnfinger]

I think the M2F bios needs to be much better optimized for high FSB and also high ram speeds, I could do much better on my P5Q-Deluxe in both respects. Disappointing considering the price of the P5Q-Deluxe compared to the M2F. Also, the M2F is way too sensitive to the vFSB:vNB voltage ratio, requiring vNB to be no more than 0.113v higher than vFSB or the board will not POST on reboot - again the P5Q-Deluxe was better and allowed you to run ~1.34v vNB and 1.11v vFSB without issues. I'm sincerely hoping the next M2F BIOS revision will address these issues.

I have never experienced this issue
I run NB volts at 1.39 and VTT at 1.33 and have had the NB volts highe than that all the way to 1.42 and it would still post and reboot.

[lowdog]

I only have to have NB V - 1.16 and VTT V - 1.16 with vcore of 1.28 LLC enabled for 4GHz - 9 X 450fsb mem 1:1. All other volts at one notch above there lowest value.

Q9650 FPO/Batch # L825B323 vid 1.2250

[CryptiK]

I have never experienced this issue
I run NB volts at 1.39 and VTT at 1.33 and have had the NB volts highe than that all the way to 1.42 and it would still post and reboot.

1.33 Vtt & 1.42 vNB = 0.09 vDiff. That's within the max of 0.113v difference I stated, so i would completely expect it to be fine at that setting. Boards may differ, but in my case, the 0.113v max vDiff between Vtt & vNB is a completely reproduceable phenomenon.

[KRich]

1.33 Vtt & 1.42 vNB = 0.09 vDiff. That's within the max of 0.113v difference I stated, so i would completely expect it to be fine at that setting. Boards may differ, but in my case, the 0.113v max vDiff between Vtt & vNB is a completely reproduceable phenomenon.

I get the same issue with my MIIF, but in my case, the maximum difference I appear to be able to set between VTT and VNB, is roughly 0.15v in the higher ranges.

Any more than that, and when I try to restart the system from Windows, the CMOS is cleared and reset and the same occurs from a cold/warm start.

As soon as I bring the difference between the VTT and VNB selection back within the permissable range, no problems at all.

[CryptiK]

KRich - Yes, that is exactly what I mean. I think the boards are too sensitive to this, we should be able to set the vFSB (Vtt) and the vNB further apart without encountering the problem you described (which is exactly what I experience too), as it is limiting the boards performance. The maximum vFSB intel allows is 1.15v and my cpu can do everything I want it to at less than this voltage. I do not want to have to overvolt my cpu simply to be able to run the vNB higher.

I could run a larger vdiff on my P5Q-Deluxe, such as 1.1v vFSB and 1.34v vNB without issues, so I know that it can be done with an Asus platform with the P45 chipset, just not on the M2F. I want Asus to address this issue, along with allowing better ram overclocking above 1200 - 1300 MHz, and increasing the boards ability to run FSB 550+ stably.

[KRich]

KRich - Yes, that is exactly what I mean. I think the boards are too sensitive to this, we should be able to set the vFSB (Vtt) and the vNB further apart without encountering the problem you described (which is exactly what I experience too), as it is limiting the boards performance. The maximum vFSB intel allows is 1.15v and mine can do everything I want it to at less than this voltage. I do not want to have to overvolt my cpu simply to be able to run the vNB higher.

I could run a larger vdiff on my P5Q-Deluxe, such as 1.1v vFSB and 1.34v vNB without issues, so I know that it can be done with an Asus platform with the P45 chipset, just not on the M2F. I want Asus to address this issue, along with allowing better ram overclocking above 1200 - 1300 MHz, and increasing the boards ability to run FSB 550+ stably.

Indeed, you are right and I agree with you.

In fact, when I first got my MIIF last week and had the problem, I was stumped as to why it was happening. It was only when I saw your first post on the subject last week that things became clearer and I was able to workaround the issue by keeping the VTT and VNB voltages within the current permissable range.

I was wondering if I needed to check the CMOS battery or whether I had installed the board incorrectly

[KRich]

As has been mentioned on this thread previously, it appears that if you run 1:1, the minimum PL you can use is 10.

Any idea if this is a P45 chipset limitation or could it be fixed via a future BIOS update?

Additionally, has anyone noticed that enabling Pull-Ins under the PL section doesn't appear to enhance the PL at all?

For example, with the Maximus Formula w/ Rampage Formula BIOS, if you set the PL to x and enabled all pull-ins, you would effectively set one PL level down from what you had entered.

This doesn't appear to happen with the MIIF, even when enabling pull-ins, the BIOS and even Everest doesn't show any change.

[ACE76]

Are people achieving 600 fsb easily or with a lot of tweaking? I simply can not get 600 with my E8600 E0 or my E8400 E0 chips whatsoever, regardless of what I set in the BIOS...but I got the DFI DK P45 and I hit it the 2nd time I played with the BIOS....is there something I'm missing here with the MIIF's BIOS?

I just noticed that DFI actually put out a .pdf guide to achieve maximum overclock with their DK P45 boards....I wish Asus would do something like that!

[ATLNY]

This is an adapted version of my setting for 4GHz with an 8400 for you to try.

Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : AUTO
CPU Ratio Control : Manual
- Ratio CMOS Setting : 9
FSB Frequency : 445
CPU Clock Skew : Normal
North Bridge Clock Skew : Normal
FSB Strap to North Bridge : Auto
DRAM Frequency: DDR2-892 MHz (ie: 1:1 ratio)
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 : 5
Row Refresh Cycle Time : 60
Write Recovery Time : 8
Read to Precharge Time : 5

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: disabled
Dram Read Training : disabled
MEM OC Charger : disabled
Ai Clock Twister : Moderate
Transaction Booster : Manual

Common Performance Level: 10 (lowest you can set is 10 on 1:1 ratio)

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 : 1.3500v
CPU PLL Voltage : 1.51325v
FSBT : 1.2v
DRAM Voltage : 2.1v
North Bridge Voltage : 1.32v
South Bridge Voltage 1.5 : 1.5
South Bridge Voltage 1.1 : 1.1

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 

Let me know how it goes. You should also know, not all e8400's do 4GHz on less than 1.45v, so vcore may be your issue.

It booted up all the way to windows, then I got the blue screen
I think it is my memory holding me back
I am going to get some corsair dominator pc8500

[CryptiK]

Try raising the vcore to 1.40v bios first, I'm not convinced its your ram. at 892MHz the ram should be ok.

@ KRich - the PL10 is the lowest you can go on 1:1 divider due to chipset limitation with P45 unfortunately. No way around it.