View Full Version : What causes prime95 to cause reboot rather than fail with error.
Devious
11-13-2007, 04:36 AM
I have got an Asus Premium P5K board (0402 bios.) and Im trying to get my Q6600 G0 (1.2125VID) small FFT stable at 4ghz.
It will run blend @ 4Ghz no problem with only 1.5v and ran 3.980Ghz with 1.52v Small FFT .
With just a small increase in FSB from 443 to 445 the pc will restart after a few seconds of small FFT's and Im trying all sorts of different bios options to fix it without success. Load temps are in the low-mid 60's before it reboots.
Now I realize this could be one of many things, motherboard, psu, cpu or memory but Im just wondering what you guys think. My psu 'should' be upto it and Ive tested the rails with a multi and they are fine.
All my ram settings are very relaxed and blend runs ok anyway so Im sort of ruling that RAM out.
I thought it may be FSB wall but I can boot it upto 515 with an x8 multi.
My vcore is relatevely low 1.52v but I tried raising it to 1.6v which you would think is enough for a 20Mhz increase but temps jump to 70+ and it just reboots the same anyway. Usually prime will error if there isn't quite enough vcore rather than reboot.
Im thinking it may be the temps, I added a couple of YS-tech fans to my RAD and it dropped the temps to just under 60 but again, same reboot.
Im thinking of trying a new motherboard, but thought I would post here before doing so. Also which board would i get? , the premium is supposed to be up there with the best.
I am happy with the lock Ive got but im so near to 4Ghz stable that Its annoying me lol.
Your thoughts...
Jon Boy Deluxe
11-13-2007, 06:04 AM
It's just not stable, I think you have reached your cpu's limit.
Mine wont boot past ~460fsb even up to 1.7v for a suicide screen shot. Iirc 455 is the best it will do for benching, prime stable 3.7ghz. I'm actually running 3.6 now with lower voltage.
Personally I'd run it @ 4.0ghz as it is blend stable... Try upping the nb voltage a touch to see if that sorts you out
Ace-a-Rue
11-13-2007, 06:33 AM
i say, be happy with 3980Ghz P95 stable at 1.52v...don't frustrate yourself any more...AND...don't change the board or anything...pat yourself on the back for what you have attained so far!:up:
celemine1Gig
11-13-2007, 08:32 AM
I'd guess that it's the board's limit.
The current draw of a Quadcore at full load at 4GHz and 1.6V is tremendous! ;)
Reboot is most likely due to voltage undershoot due to too much current draw.
With one of the high end boards with a higher rated PWM, you might be able to get it stable. But as already said, I wouldn't try. Could also be that you won't be able to get even the same overclock on another board.
Devious
11-13-2007, 08:39 AM
Thanks, ill keep trying, its just annoying being so close to 4Ghz, it was exactly the same with my E6600 though i had run out vcore with that. There seems a lot more in this chip its just a shame i cant squeeze it out.
With one of the high end boards with a higher rated PWM, you might be able to get it stable. But as already said, I wouldn't try. Could also be that you won't be able to get even the same overclock on another board.
Any board in mind? I may be able to borrow on ;)
celemine1Gig
11-13-2007, 08:47 AM
Perhaps a Gigabyte P35 or X38 DQ6 with 12-phase power, a DFI P35 Lanparty or a Foxconn Mars. But I'm only guessing. Never had one of these. You'll most likely have to ask some people with more experience on the really xtreme overclocks. I'm more into modifying the boards and then never really exploring the real maximum due to limited time. :D
suffocatio
11-13-2007, 10:06 AM
CPU VOLTAGE REFERENCE rise or lower.
celemine1Gig
11-13-2007, 10:22 AM
I doubt that this will help to get rid of the reboots, but give it a try anyway. ;)
Devious
11-13-2007, 10:48 AM
I did previously find lowering FSB termination to it lowest helped get above 3.9Ghz but i cant go any lower now.
Devious
11-13-2007, 10:49 AM
CPU VOLTAGE REFERENCE rise or lower.
Tried all valuies for that and auto with no luck.
camouflage
11-13-2007, 01:45 PM
;) Same here on P5E - but it's stable, I ran more than 3 hours prime test with the Hot CPU Tester.
Somewhere I read that Prime95 & Orthos has problems to run on some Intel boards:
http://img80.imageshack.us/img80/3298/4ghzprimewj8.jpg
Orthos caused a fatal crash each time started at the same settings....:stick:
[XC] DragonOrta
11-13-2007, 03:19 PM
Did you try Northbridge Reference voltage? I was getting reboots on my P5K-E/Q6600 until I set it to 67x.
Did you ever resolve this m8??
I've got an L737B (Vid-1.225v) Q6600 cpu here and i get reboots with P95 small FFT's at 3.8Ghz regardless of vcore on my P5K-Premium !!
My L726A Q6600 runs fine at the same settings and has a higher vid of 1.2875v.
Devious
01-03-2008, 04:38 AM
It was all down to temps m8, this chip is very sensitive to heat. The higher i went over 3.8Ghz the cooler it had to be. At 4.0Ghz it would reboot when running Smal FFT if a core hit 52c . Hence i had upgraded my cooling and it never went over 50/51... phew ;)
Mobo settings made no difference, they are all pretty much back on auto now except for the obvious.
Bail_w
01-03-2008, 04:44 AM
i dont think it has to do with CPU temperature. Example, when i put 1.6v on my quad, its (prime95) load temperature were around 55 but it reboot my system. on the other hand, when i put 1.6v on my other quad, its load temperature were around 70 and it still reboot my system. Both were under my watercooling loop. Some quad G0 just runs cooler than others.
Devious
01-03-2008, 04:49 AM
i dont think it has to do with CPU temperature. Example, when i put 1.6v on my quad, its (prime95) load temperature were around 55 but it reboot my system. on the other hand, when i put 1.6v on my other quad, its load temperature were around 70 and it still reboot my system. Both were under my watercooling loop. Some quad G0 just runs cooler than others.
Yes, but you were talking about two different quads, one is more sensitive than the other. I was saying my particular chip is also sensitive.
I see people on air running prime95 with temps over 70c at 3.8Ghz with 1.65v, no way would my chip do that. But theres will probably never do 4ghz, even if they had top watercooling. Miine does 4ghz wit just 1.5v.
The L726A's seem to be very stable with high temps.
Bail_w
01-03-2008, 05:38 AM
I would say it has to do with the motherboard.
vBulletin® v3.7.0, Copyright ©2000-2008, Jelsoft Enterprises Ltd.