ASUS Maximus II Formula - new P45 king?

Quote:

---

Originally Posted by CryptiK

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.

Quote:

---

Originally Posted by In-Fluence

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.

---

if your failing large FFT's 1024k test then you need more VTT or vNB

1024 - increase VTT or vNB
896 or 448 - vNB
768, 512, 332 - VTT

Yes I'm aware of this, but it passed 1024k every time when I was blend testing with different CPU GTL combinations (that was over 25 times and always failed during the first two small FFT tests). I then leave it off over the weekend, come back and run the test with the exact same settings, volts etc, and the most succesful CPU GTL setting (so far) - then 1024k fails within 5-10 minutes. This is why it's so hard for me to tune this thing - it's simply not playing ball the way it should.
I've tried another restart and power cycle, and it's still failing at the same point - I've wasted so much time with this I'm tempted to at least RMA the board provided Arseus allow it, failing that I could hope that my new PSU (due to arive soon) could help the matter. If not I'll probably sell the board and try a gigabyte on for size. Sorry about these constant rantings, but It's really bugging me! I'm no noob to overclocking (since about 2000 and mainly spent time on the OCW forums) and I feel my experience is often enough to work out which settings need tweaking to get stable and get a good feel for how a motherboard operates.

Sorry, this isn't a dig at anyone taking the time to help me, it's always greatly appreciated, but I hope you see where I'm coming from! :)

In-Fluence, use small FFT to determine cpu voltage, large for NB and VTT as Mr. Pink listed and then go to blend. This worked like a charm on my Xeon...not so much with my 9505 (which i still apparently don't have 100% stable for WCG)

That's exactly what I have been doing when finding a stable spot. I tuned the vcore using small fft for about 4 complete loops. I then moved onto the NB/memory and tested with large fft for another 4 loops, then ran 25 large linpack, followed by 12 hours of blend, so I can quite easily get stable as far as I'm concerned.
It's just the GTL tuning which I really thought I was getting somewhere with until I returned after this weekend and powered up with the same settings.

Like I've said, I got better results with -ve GTLs after I lowered vcore to make it unstable and blended. When testing and comparing -55,-5,-55,-5 (and upwards) to +50,+10,+50,+10 (and many combinations in both directions) the -ve ones (with a 50mv gap) gave me an average of 25 minutes of extra stability (well into 8k, and often into 12k, as mentioned before) which made sense to me (even though it seemed a bit odd given the VTT of 1.32, the 50mv gap, and the equation). When testing the +ve GTLs I got blend failures within 10 minutes of 1024 which indicated much more instability.

If I up the vcore to one notch below a completely stable vcore (1.45v when at 3.6GHz) and try to find the GTLs when I'm seemingly 98%+ stable already (i.e able to prime for hours before failure/reset) this is clearly going to take far too long. The trickiest thing is the fact it keeps rebooting which although seems like a sign of greater instability, doesn't help when you have to keep staring at your screen ever 30 seconds to find out how long it lasted or look through log files if you miss it!

The only reason I went back to tuning the GTLs after being 100% stable is because I read a lot about the GTLs helping to lower vcore if tuned right (although I understand this isn't always the case).
After discovering that -ve GTLs worked better and gave me 8K stability with 1.425vcore (I needed 1.44v with +ve) I thought this was worth persisting with, but after returning this weekend, it all went out of the window!

I'll wait for my PSU and see how things go from there and probably make a decision about what to do. The inconsistencies I'm experiencing is now the main problem I'm having, nothing to do with getting stable, I just want to tune it properly!

Quote:

---

Originally Posted by In-Fluence

The inconsistencies I'm experiencing is now the main problem I'm having, nothing to do with getting stable, I just want to tune it properly!

---

understandable. Personally, after my experience with tuning the GTLs i don't think the board is made right or the equation they provide is completely wrong since the equation gave me -5,-45,-5,-45 and i ended up with 90,50,90,50 as being the most stable

Quote:

---

Originally Posted by In-Fluence

Yes I'm aware of this, but it passed 1024k every time when I was blend testing with different CPU GTL combinations (that was over 25 times and always failed during the first two small FFT tests). I then leave it off over the weekend, come back and run the test with the exact same settings, volts etc, and the most succesful CPU GTL setting (so far) - then 1024k fails within 5-10 minutes. This is why it's so hard for me to tune this thing - it's simply not playing ball the way it should.
I've tried another restart and power cycle, and it's still failing at the same point - I've wasted so much time with this I'm tempted to at least RMA the board provided Arseus allow it, failing that I could hope that my new PSU (due to arive soon) could help the matter. If not I'll probably sell the board and try a gigabyte on for size. Sorry about these constant rantings, but It's really bugging me! I'm no noob to overclocking (since about 2000 and mainly spent time on the OCW forums) and I feel my experience is often enough to work out which settings need tweaking to get stable and get a good feel for how a motherboard operates.

Sorry, this isn't a dig at anyone taking the time to help me, it's always greatly appreciated, but I hope you see where I'm coming from! :)

---

Unless your FSB is above 450 GTL's are not really a factor for this board.
I too overclock and I do it rather well, I have been here for along time dont let the account fool you.
I think your letting your frustration get the best of you..have a pint relax and start from scratch..
I played with many bios versions I find some clock better than others for an example my 8400 clocks better on bios 1802 while my 9650 clocks better on 1307.. I own 2 MIIF boards and both run/clock very well. Its possible you have a faulty board but Asus will ask if its stable at stock

I have been searching everywhere on this thread for a Similar setup as mine i have tried all sorts of ways to keep it stabled anything passed 3.5Ghz can someone send me a layout that can help me it will be greatly appreciated.

Asus M2F
q9550 45NM yorkfield.
8GB PC1066 GSkill
800PSU
Water Cooling on CPU and NorthBridge
Ati 4800 Crossfire.

Right now im back to default this has been so frustrating. I was able to overclock my 6600GO on a P5WDH with very little Fuss.

Thanks in advanced.

You're quite right Mr.Pink. I'm definitely going to take a step back, chill, and have a good think. I'll definitely try 1802 as well (I obviously missed the fact that this was actually a good one). I've tried 2202, 2104, 1901 and 1307 without much luck, but who know, maybe 1807 or any lower ones might help (if anyone knows which ones are good, please let me know).
Like you also say if gtls definitely need only be tuned whilst higher than 450, then I guess I could do without them. I'm just eager to get past this frustrating wall of 3.6GHz when I know I can get better out of it. >400x9 is an option I have also seemingly exhausted, so more bios are probably my last resort. 1307 seems to have done best for me so far.

...

Just a quick update - tried 1802 and managed to boot at 490fsb (still only 2 cores) which I couldn't do on any other bios, plus copy scores seem much better now. Still not easily stable, but this one seems better (so far).

I'm running bios 1901 and am able to 485 fsb stable in prime blend overnight but highest clockspeed thus far for my q6600 is 4.1ghz hit 4.2 on my asus P5Q-premium weird the rog board wont clock as high as the mainstream board did

One further update, just tested the battery with a dmm, and it was about .3v less than a fresh one, so I put that in and have done a full clear of CMOS, perhaps this will help with the inconsistencies, perhaps bios settings werent getting saved properly?...

Just got my corsair HX850 too - 12v and 5v are above spec (with other psu they were lower), and vcore is being reported nearer to actual bios setting now - fingers crossed! :)

Quote:

---

Originally Posted by overclocking101

I'm running bios 1901 and am able to 485 fsb stable in prime blend overnight but highest clockspeed thus far for my q6600 is 4.1ghz hit 4.2 on my asus P5Q-premium weird the rog board wont clock as high as the mainstream board did

---

nice :up:
maybe you could post your profile as to help others with similar chips

whats up at the MIIF frontline?
what about that new 2202 BIOS? does it perform better?

Not a great deal mate, seems the 2202 doesn't offer a great deal more. I'm not really having a huge amount of luck though :/

Does anyone find they have to use different GTLs for different bios or can they be the same?

Doesnt seem like the new PSU has done a lot for me :/

Quote:

---

Originally Posted by Mr.Pink

nice :up:
maybe you could post your profile as to help others with similar chips

---

sure thing m8 ;) is there like a template to use? i'll just post it rough on the settings I use right now if a setting is not listed then it's left on auto.

overclocking: Manual
Ratio: 9
FSB: 467
Strap: Auto
cpu delay: 200ps/300ps
nb delay: auto
all mem timings: auto
all mem tweakings: set to best for overclocking or auto
PCI-E: 105
CPUV: 1.61250v
CPU PLL: 1.80V
DRAMV: 1.850V
FSB: 1.475v
NB: 1.430
SB 1.5: auto
SB 1.1: auto
CPU GTL 1: +70 2: +40 3: +70 4: +40
NB GTL: +50
Spread spectrum: disabled
LLC: enabled

this is just for raw fsb and high clock stability, i have another that the memory settings are better with higher mem clocks but I was not able to achieve 4.2GHZ with the better memory serttings.

My setup:
Q6600 VID=1.2v
m2f
4GB OCZ low voltage 1066mhz
2XHD4870
ultra 1000w psu
all watercooled nb water cooled to
vista X64

any other questions?

Wow, that's a lotta volts even for WC. I assume that's just for benching? What are your CPU temps like?

Wow, I am new here, and just finished reading this thread.
Looks like only one guy managed to master this board and run it at 500FSB 24/7 with a quad. There were some few people showing their quads at 490-500fsb, but without any stability tests, so that means nothing.

Here is simps's quad, and its pretty much the only one I found on this entire thread, running 500fsb 100% stable.

494FSB (wtf, his board volts are so low for that fsb)
http://img297.imageshack.us/img297/5383/4945ju6.jpg


500FSB (the only quad 500fsb stable overclock on this entire thread)
http://img80.imageshack.us/img80/5244/new50013wu3.jpg

He is aircooled he said, that is really impressive. I can't make my quad 500fsb stable on this board, even with lower cpu clocks.

Quote:

---

Originally Posted by In-Fluence

Wow, that's a lotta volts even for WC. I assume that's just for benching? What are your CPU temps like?

---

yep those are high volts but 4.2GHZ settings. just got it stable last night, in order to get 485 fsb i have to drop the multi to 8 settings are the same ecxept vcore is a bit lower thats it will post at 492 fsb but as soon as it hits above 40% load it crashes

Quote:

---

Originally Posted by Marcio

Wow, I am new here, and just finished reading this thread.
Looks like only one guy managed to master this board and run it at 500FSB 24/7 with a quad. There were some few people showing their quads at 490-500fsb, but without any stability tests, so that means nothing.

Here is simps's quad, and its pretty much the only one I found on this entire thread, running 500fsb 100% stable.

494FSB (wtf, his board volts are so low for that fsb)
http://img297.imageshack.us/img297/5383/4945ju6.jpg

He is aircooled he said, that is really impressive. I can't make my quad 500fsb stable on this board, even with lower cpu clocks.

---

seems he got lucky and got a really good board, don't know if it helps that he is running an extreme edition cpu.

with my Xeon x3330, i was able to boot at 490fsb with 1.38 vtt and vnb and 1.40 vcore on the 333 strap, i had never messed with the strap until this week so i might be able to get stable at 480+ if i try. i may look into it after i try the chip on a UD3P

Quote:

---

Originally Posted by overclocking101

yep those are high volts but 4.2GHZ settings. just got it stable last night, in order to get 485 fsb i have to drop the multi to 8 settings are the same ecxept vcore is a bit lower thats it will post at 492 fsb but as soon as it hits above 40% load it crashes

---

Stable with 1.6v 4.2GHz and on watercooling? Or do you just mean bench stable? Let's see some screenies then!

I managed 4.08GHz which was just about stable enough for a 1M superpi run, and 4.05GHz for a 32M run, but didn't mess with it too much as I'm still trying to get a 24/7 stable config, but may try for more soon :D

Gone back to 1901 after having multiplier problems with 1802. Stabilising the system with the same settings used on other bios wasn't successful. Does anyone have to use majorly different settings for different bios (in particular GTLs, skews & voltages)?

Quote:

---

Originally Posted by In-Fluence

Gone back to 1901 after having multiplier problems with 1802. Stabilising the system with the same settings used on other bios wasn't successful. Does anyone have to use majorly different settings for different bios (in particular GTLs, skews & voltages)?

---

hmm, that could explain things with my issues OCing the 9505, guess i'll have to try 1901

Quote:

---

Originally Posted by Marcio

Wow, I am new here, and just finished reading this thread.
Looks like only one guy managed to master this board and run it at 500FSB 24/7 with a quad. There were some few people showing their quads at 490-500fsb, but without any stability tests, so that means nothing.








He is aircooled he said, that is really impressive. I can't make my quad 500fsb stable on this board, even with lower cpu clocks.

---

you cant compare 45nm to 65nm on this board, might as well compare dualies to quads:rolleyes:

45nm should be easier to OC but I am beginning to wonder now.

Asus are lamers, lack of support is main issue with this Mobo.

I might pop in my old 6600 for fun it did 3.8GHz stable on X38 and could hit 4.0GHz bench stable, and would do 500FSB altho I had to disable 2 cores to do it

the settings are prime stable 4 hours but with ibt it overheats and shuts down. I'll post screenies soon enough I have lowerd it back down with the multi trying for 500fsb with this quad but it just wont do it