ASUS Maximus II Formula - new P45 king?

[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

[Haes]

Flex was faster ...

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

[humeyboy]

WOW, What have Asus f00ked now.

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

[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 (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 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.

[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]

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

[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.