Modded ASUS P5Q/Pro/Deluxe/Premium BIOS

[SPiTFiREgr]

Well I have tried ANY settings combination that's available. That was ~ 3 weeks of testing and stressing. The conclusion i came is that when writing bandwidth excess a certain amount the system is unstable. That would be either a chipset fault or a memory incompatibility (?) Unfortunately i couldn't find any other P5Q-PRO owner with the sticks i tested (SUPERTALENT 800Mhz CL4, OCZ SLI EDITION 1066 CL5) to have a reference. Everyone seems to have corsair or g.skill or crucial..

I have contacted ASUS support and co-operated on researching the problem and they said me that maybe it's a Northbridge heat problem..

Also i have to mention that when i say "system unstable" i mean Prime95 blend. When using small ffts test there is no instability.. So that tells me that it's something to do with the memory. And because of the high fsb even lowering the memory to 1:1 and running it at 800Mhz doesn't lowers memory writing bandwidth..

[Ket]

I have a kit of that OCZ SLI memory, maybe we can figure something out.

[samsonJS]

MajkY: You're not alone. A couple of us in the P5Q thread on [h]ard have the same problem, though mine isn't quite that bad. The only advice I can give is to reflash your BIOS, clear CMOS, load default settings, and definitely do not load any saved OC profiles, create fresh ones. After doing all that sometimes the problem returns after 4-7 days but for me it's mostly fixed. Now and then it crops up, maybe 1/50 or 1/100 times I save BIOS settings. I really want to know what triggers it as it seems random on my board.

[SPiTFiREgr]

I have a kit of that OCZ SLI memory, maybe we can figure something out.

The part number of the sticks is OCZ2N1066SR1G and they are 2 x 1GB kit . Do we have the same? Have you ever tested them with the P5Q PRO?

Oh man you maybe be the first person i meet on the net to have a P5Q PRO and some same memory as mine. lol I had started to think that i would have better made a research for what memory most P5Q PRO owners use before i made my purchase. In fact i didn't know that G.skill or Crucial was so famous.. I thought i would find more people using OCZ memory

[Dashock]

First off Ket of to you my good man. Second i got all the settings on auto at the moment i know youve said it might be better leaving them loose but do you think i should try to you know configure them again and take them off auto might increase in performance ( might ). I really wanna do 5-5-5-12 on these Mushkin Ascent Redline 8000 kit i got it running at 1066.

[ahmad]

I ´m so pissed offfffffffffffffffffffffffff

when i press start-restart from windows, system restarts, but thats all. No boot at all...... no logo screen nothing. (fans are on) and this is happening everytime i want to restart system no matter if its from windows or bios... also when i change some setting in bios "save and exit" from bios...it restart, but no boot... i have to unplug electricity and than on 2-3 time it boot.
I have ASUS P5Q PRO with bios 1004. I wonder if this 1307 for my board can solve this problem ? :S

This is a voltage issue I am sure. Try higher vtt..

[Ket]

The part number of the sticks is OCZ2N1066SR1G and they are 2 x 1GB kit . Do we have the same? Have you ever tested them with the P5Q PRO?

Oh man you maybe be the first person i meet on the net to have a P5Q PRO and some same memory as mine. lol I had started to think that i would have better made a research for what memory most P5Q PRO owners use before i made my purchase. In fact i didn't know that G.skill or Crucial was so famous.. I thought i would find more people using OCZ memory

Yep thats the same OCZ kit as I have. I wasn't looking for memory that clocked really well, just something that was fast stock and maybe had enough wiggle room to go to ~1150MHz. After 5 RAM kits I'm convinced the memory I have can't all be crap clockers, and theres another limiting factor somewhere.

First off Ket of to you my good man. Second i got all the settings on auto at the moment i know youve said it might be better leaving them loose but do you think i should try to you know configure them again and take them off auto might increase in performance ( might ). I really wanna do 5-5-5-12 on these Mushkin Ascent Redline 8000 kit i got it running at 1066.

I set some timings manually, but for the most part other timings are best left at auto anyway.

[MajkY]

This is a voltage issue I am sure. Try higher vtt..

I couldn´t belive myself...but you are right. i set 1.4V VTT and no reboot issue !
Thank you very much !!

- anyway i will try new bios 1307

[carepolice]

Majky 1.4 VTT is maybe too high for 45nm chipset !!

[MajkY]

I didn´t know. 1.32VTT should be OK ?

[carepolice]

By myself max VTT that i use is 1.34V, Many people says that ~1.35 VTT is fine. Try up some other voltage, But remeber: stay in safe values :>

[SPiTFiREgr]

Yep thats the same OCZ kit as I have. I wasn't looking for memory that clocked really well, just something that was fast stock and maybe had enough wiggle room to go to ~1150MHz. After 5 RAM kits I'm convinced the memory I have can't all be crap clockers, and theres another limiting factor somewhere.

Do you mean that you have tested OCZ kits with the P5Q PRO and there is a limiting factor somewhere else but the sticks themselves? When you say 5 ram kits you mean five OCZ2N1066SR1G kits?

I'm currently running those OCZ2N1066SR1G at 1154Mhz with 360FSB (5:8 ratio) and they are stable. But as i said raising the fsb (and as a result, memory write bandwidth) there is instability with blend stress (large fft).. Do you mean the same?

If high memory write bandwidth is the cause of the instability with certain memory modules then the only thing i can think of as the cause of the problem is inappropriate timing settings set by the motherboard (because there are some timing options that cant be set by the bios. even with manual selected ofcourse).

[bayer]

It's definitely a voltage problem, or at least a VTT issue. I had the SAME issue as a lot of people where the board would boot fine the fist time, but then fail to boot if restarted (via Windows). For me, I was afraid to go above 1.2 VTT with my E8500 but decided to try it out and see what happened... I set VTT to 1.22 and booted the board, and then restarted via Windows and it actually restarted this time without issue. I then kept lowering VTT until I was satisfied with the VTT value and the stability of my OC. Right now, I've settled for VTT at 1.16 and the board boots and restarts fine.

The funny thing is though, that I never had this problem when I first bought the board... It happened about 2 weeks after, and lasted for a few days. Loading new BIOS's and even a full clear of CMOS with the battery out didn't seem the help. Then one day, just setting VTT above 1.2 got the board acting normal again.

[xdan]

The fact is that many P45 series have a strange comportament, in special from Asus. The strange things is that many work normal, others don't, the bioses don't use in my opinion the full potential of chipset, i start to think that
all these ar because all mb were made in a rush like "hotyes 4850" just to take money from us. And the most condamnble thing is that they even bother to repair in time the bugs with bioses, and without men like Ket many of us would have even now issues.

[SPiTFiREgr]

It's definitely a voltage problem, or at least a VTT issue.

I have tried every voltage available. I have even tried NB as high as 1.42 and VTT as high as 1.38

45nm chips do not need so high fsb voltages to go for 400Mhz FSB..

I have also tried higher fsb to exclude any possibility of fsb hole. I have tried 401, 403, 405, 410, 415, 420.. All with the same result!

[Vazovskiiii]

I think they are still beta yeah. I'm using a modded 1307 on my Pro and its working fine, from what others have said the other betas are fine as well.

ket do you have 1307 your mod bios for P5Q non pro? can you arange this for me ?

[Sauron_Vie]

@Ket: the m0803 Bios from ur first post is also for the p5q deluxe ?

[Vazovskiiii]

here is my overclocking on 1306 bios


waiting for 1307 from ket

[ThugsRook]

waiting for 1307 from ket

check the main post, its been there for awhile now.

[hema]

here is my overclocking on 1306 bios


waiting for 1307 from ket

here

[SPiTFiREgr]

Another fact: Trying to set 400 FSB with 333 STRAP and 960 Memory Speed resulted in post fail with beep beep beep beep beeps (not beeeeeeeeeep, many beeps lol)

Ket, also tried your 1004 modded bios

[Dashock]

Alright i know some of you have been having issues with Load Line calibration turned on/off i currently have it on Auto. I havent seen any of these issues i got it up to 4.5ghz i was only 2 hours stable. But i only had it pumped to 1.41volts so a bit more voltage im pretty sure i can get it stable but i would not be running that 24/7 i just wanted to hit 4.0ghz and i achieved that. My motherboard might be a new revision Asus P5Q-E ReV 2

[Ket]

Scout your board for the rev, if you have a newer rev it'll be higher than 1.03G.

[carepolice]

What changes is there in newer versions of rev? CPU-Z says for me rev:A2 is that old or new?

[Ket]

Do you mean that you have tested OCZ kits with the P5Q PRO and there is a limiting factor somewhere else but the sticks themselves? When you say 5 ram kits you mean five OCZ2N1066SR1G kits?

I'm currently running those OCZ2N1066SR1G at 1154Mhz with 360FSB (5:8 ratio) and they are stable. But as i said raising the fsb (and as a result, memory write bandwidth) there is instability with blend stress (large fft).. Do you mean the same?

If high memory write bandwidth is the cause of the instability with certain memory modules then the only thing i can think of as the cause of the problem is inappropriate timing settings set by the motherboard (because there are some timing options that cant be set by the bios. even with manual selected ofcourse).

Not 5 OCZ kits no. I thought the P5Q series was just very picky about what RAM it would play nice with, thats seems to not be the case tho I have tried to date;

Patriot PC6400 4-4-4-12 (works great @ 1066 on a AW9D-Max) on the P5Q series the system won't even POST with this kit.

Geil Golden Dragon PC6400, I'm sure on any other board this kit would do 1066+, on the P5Q series it boots all the way up to 1100+ but isnt stable even @ 1029

OCZ SLi kit, constant problems with the memory, booted 1143MHz, not stable. Again I think this is a P5Q series BIOS issue.

Mushkin PC8000 D9GMX, booted 1143MHz, not stable.. (noticing a pattern here yet?)

Cellshock PC8000 D9GKX, boots up to 1206MHz, not stable even @ 1143MHz seemingly..

The only explanation to all this I can think of is my CPU and the P45 not actually being properly tailored for 65nm CPUs.. that or asus partially fudged 65nm support. Even on a AW9D-Max I can run my "lowly" E4400 on a 400 strap @ 425FSB, absolutely perfect. What happens when I try setting 400 strap and memory divider on the P5Q boards? No POST. Well done Asus /sarcasm.

The timings you can't set in BIOS are not a huge problem, this is where spdtool comes to the rescue, you can loosen things like trc via the spd data and reflash the modules. The only pain in the arse is that w1z still hasn't added P45 support. Bad w1z, bad bad w1z

Theoretically loosening trc for intel boards should allow for better stability and higher frequencies.