Overclock won't pass Prime95

I've had Prime95 run OK one moment, then crash after a few minutes the next. Same overclock. I've never had my hard drive corrupt due to an overclock, and I really do think it depends on what you do with your computer.
 
Qwertyman said:
ERROR: ILLEGAL SUMOUT
Possible hardware failure, consult readme.txt file, restarting test.

you people dont take a hint when you start getting errors like this do you? when your processor cant add numbers correctly theres a problem.
 
acascianelli said:
you people dont take a hint when you start getting errors like this do you? when your processor cant add numbers correctly theres a problem.

well duh i know there's a problem thats why i'm asking this board for help to fix, 1) should i clock back down to stop, 2) should i leave it as it even though i do get runtime errors on UT2k4 if i play for a while because it doesn't do Math. 3) should i tweak my settings a little more i'm at 193x11 at 1.775v on a MSI board with the latest bios. Thats why i'm asking for your help not for an argument because i know its not stable as to reason i wanted to run prime to test.
 
If it's giving you problems playing games, and not just in a torture test, then you do have a problem. The first thing I would do if you don't want to lose the clock speed would be play with the settings. If you don't want the aggravation, you could just downclock it a bit till you get stability back.
 
Qwertyman said:
well duh i know there's a problem thats why i'm asking this board for help to fix, 1) should i clock back down to stop, 2) should i leave it as it even though i do get runtime errors on UT2k4 if i play for a while because it doesn't do Math. 3) should i tweak my settings a little more i'm at 193x11 at 1.775v on a MSI board with the latest bios. Thats why i'm asking for your help not for an argument because i know its not stable as to reason i wanted to run prime to test.

sorry, i wasnt trying to criticize you for not knowing there was a problem, i was just using your post with the errors from prime95 to show other people that when you get errors like that, that your cpu is not stable.

i was directing it towards people who say that prime95 stability isnt important.
 
I need a completely stable system because I fold, so I am in the camp that believes that if your comp doesn't pass a long run of Prime95, then your comp is unstable. If you feel that Prime95 isn't cutting it, might I suggest CPUBurn[/URL[? It's a program that is designed to maximize the heat output of the CPU while also stressing it with constant calculations. If your CPU is faulty or 'weak,' it'll probably fail in the first 10 minutes.

@poster before me, I think the 2500+ desktop version has a default VCore of 1.65 V.
 
Qwertyman said:
[Mon Jun 14 08:25:56 2004]
Self-test 8K passed!
FATAL ERROR: Rounding was 0.4999694824, expected less than 0.4
Hardware failure detected, consult stress.txt file.

I get that same error when testing with the max heat test. Does them mean i should trying upping VCore? If so.. to what? right now my Vcore is set to 1.65V, but My mobo Asus Probe reads 1.696V

what should i do?
 
well here's whats working for me SO far. i have yet to do an overnight or 24hr test but i am now at 190x11 with a 1.75vcore and it is running stable(it passed the tests i've failed so far, i'll up the fsb later and try again some other time), try that. also if anyone knows or has done an overclock on cpu stepping code AQZFQ write back your results because this is a fairly new chip, well i just got it from newegg about a month ago.
 
OKay.. i think this overclock damaged my CPU fan.. i have a silent boost.. after running Prime for 45 minutes.. i get a warning that my CPU fan is under 600.. my CPU fan should be at about 2500!!! what's going on? i put the comp back to stock speed, and my temps are higher than before >=\

i'm confused.. i don't know if it's always been like that? or what... i'm really confused.. my first overclock is looking scary now.

EDIT:

Now that i look at it again.. i think my mobo fan speed reader is messed up... i mean i have this external fan speed reader and it tells me my fan is fine.. what's the deal?
 
The overclock seems to have screwd up my fan readings... is this normal? i restarted (i'm at stock speeds now) and everything looks normal again... any clues?
 
no idea i'm assuming u have everything plugged in right? i have the silent boost and no problems so far from my perspective.
 
yea everything is plugged in perfect... but yeah.. i don't know.. you think i should try upping the FSB again? im thinking of going to 180 this time... before it was at 200 with Vcore = 1.75V it was working great except for that fan thing.. The temps were good to.. it was at 56c after 45 minutes of the max heat test in prime. BUt like i said.. that's when that weird fan thing happened.
 
Those weird fan readings popped out again (the fans don't seem to slow down.. i mean if they did my case should get very quiet... is there anything else i should check in the BIOS.. like RAM and stuff.. or should i just give up on the OC? i don't know.. it's sort of a all or nothing attitude i have.. if not 200x11 then 166x11 nothing in between... that's what i want.. grrr lol...
 
God I hope none of you guys that think you dont need to be Prime95 stable are folding, searching for ET or a some drug.
 
CentronMe said:
God I hope none of you guys that think you dont need to be Prime95 stable are folding, searching for ET or a some drug.

It mentions that in the folding readme, and i would say there no reason to argue. Gamers however, are a whole differant animal.
 
I figured out the Fan situation.. there seems to be some software conflict between ASUS prob (my on board temp,voltage, and fan speed reader program) and CPU-z. I noticed that the problem would occur when i run CPU-z for a while at the same time with my ASUS probe.. the FAn speed problem even happened when i was running stock speed and voltage!!... so i'm gonna OC again and test the hell out of it with Prime, making sure i don't run ASus probe and CPU-z at the same time.. i'll report if I finally get stabillity with prime..

thanks
 
MemoryInAGarden said:
I'm not a tremendous believer in Prime. I've had this chip pass hours of Prime at 3.2 ghz 1:1, and it wouldn't play 15 mins of Quake 3 without a CTD. Now 3.15 1:1 is both prime and game stable, so I use that as my day-to-day clock speed. While I can run 3.25 5:4 prime and game stable, the 3.15 1:1 generally gives better performance.

There is a reason for this. Prim95 only tests memory / cpu. Well if you are close to the limit of your current clock / voltage, and you start gaming and your video card adds a lot of heat to your system, this can cause your system to have problems, so keep this in mind.
 
So far i have been running prime95 for about 16 hours with no problem on the barton 2500+. I have it at 200x11 and Vcore is = 1.700V ,, but my asus probe is reading 1.744V. My temps are looking good.. i tried to make the room as hot as possible and i got a reading of 56C. right now with the room at it's normal temp it reads about 53-54c... what do you guys think?

I'm gonna keep running prime till about 1:00 AM that will complete alittle over 24 Hrs of testing... you think i can call this a sucessful overclock? This is my first time over clock :)
 
u know what i fricking hate u, my crappy cpu won't do garbage. what's your ram timings at cause mine still fails at 4min on 32FFT. and what did u bump your voltage up to?
 
I ran my xp2500's @ 2.0V on air and they ran for well over 6months before they were sold and after that they are still out there running. Give that bitch some juice or just give up.
 
I'm not sure what my RAM timings are.. i left them alone.. though i did set my system performance to "aggresive" in the BIOS.. whatever that means... as for voltage i set it to 1.700V but when i check the voltage with the on board monitor it say's 1.744V
 
crap at 18 horus and 10 minutes it failed!! >=\.... any suggestions?

Should i raise Vcore to 1.75V? i had it at 1.700V.. wht do you guys think?
 
i'm at 1.775v at 190 and its working well SO far. running a 32FFT test again ram timings are 2-2-3-7
 
^^ I don't remember where i got it.. i think you just type Prime95 in google....

OK here's a little problem i'm having... i'm currently back to stock speeds and voltages... The problem is.. i have a fan that's powered through the motherboard.. but my ASUS probe shows me that the RPMs jumps between 3500 and like 6000 RPM i do'nt know why it's doing that, so before i try to over clock again i want to fix this problem.. should i just connect that fan with a molex directly to the power supply? If i do this i won't be able to monitor the fan anymore... what do you guys think?
 
PhattyT said:
ummm im a noob and was wondering where yall dl P95? :D

http://www.mersenne.org/freesoft.htm

Prime95 is thine overclocking God. Thou shalt worship it and obey it. Let no false idol or vile naysayer distract thee from the One True Faith.



Seriously, your PC works or it doesn't. It can do sums or it can't. What are we arguing about again?
 
well good news i failed it again. (sarcasm) Small FFT's Tortune Test here's my results at 1.775 190x11 2-3-2-7, i hate this stupid mobo:
[Thu Jun 17 06:25:19 2004]
Self-test 8K passed!
[Thu Jun 17 06:44:00 2004]
Self-test 10K passed!
[Thu Jun 17 07:01:22 2004]
Self-test 12K passed!
[Thu Jun 17 07:19:46 2004]
Self-test 14K passed!
[Thu Jun 17 07:38:56 2004]
Self-test 16K passed!
[Thu Jun 17 07:54:15 2004]
Self-test 20K passed!
[Thu Jun 17 08:10:57 2004]
FATAL ERROR: Rounding was 0.4999656677, expected less than 0.4
Hardware failure detected, consult stress.txt file.
 
no it blows, its a 2500+ not a 2600 its wrong in my sign i will change later though. i don't know why its being a bitch i've been try to get a stable overclock for about 3 weeks already. MSI YOU SUCK!
 
I don't want to try and OC this chip again till i resolve this fan speed issue... but uumm when i start OCing again.. if this thread is alive.. i will post up results..
 
Qwertyman said:
no it blows, its a 2500+ not a 2600 its wrong in my sign i will change later though. i don't know why its being a bitch i've been try to get a stable overclock for about 3 weeks already. MSI YOU SUCK!

try messing with your memory timings, try setting them to like 2-3-3-8 or even 2.5-3-3-8 if the others dont work. and you can always bump the voltage up some more. i have a 2500xp also in my other computer and i was able to bump it up to 210x11 at around 1.8 or 1.9v. i dont remember right now. but try some different memory timings and see if that does anything for ya. good luck dude.
 
Ok.. seems like the fan has a screwy yellow signal wire.. all other fans work great on the same fan connector...I'm probably just gonna replace the fan sometime soon so back to OCing!!! but before i do..

These are the RAM timings i'm getting from the BIOS..

SDRAm active Precharge delay = 8
RAS to CAS delay = 4
SDRAM RAS precharge delay =4
CAS latency = 2.5T

would this timing read as 8-4-4-2.5T ??? in what order do you label these things.. also.. do these timings sound reasonable for the overclock i'm trying to get to..

This time I'm just gonna up the voltage to 1.75.. i was on 1.700 when it failed prime after 18 hours and 10... let's see what happens.. i will post my result
 
RancidWAnnaRIot said:
Ok.. seems like the fan has a screwy yellow signal wire.. all other fans work great on the same fan connector...I'm probably just gonna replace the fan sometime soon so back to OCing!!! but before i do..

These are the RAM timings i'm getting from the BIOS..

SDRAm active Precharge delay = 8
RAS to CAS delay = 4
SDRAM RAS precharge delay =4
CAS latency = 2.5T

would this timing read as 8-4-4-2.5T ??? in what order do you label these things.. also.. do these timings sound reasonable for the overclock i'm trying to get to..

This time I'm just gonna up the voltage to 1.75.. i was on 1.700 when it failed prime after 18 hours and 10... let's see what happens.. i will post my result

nah most folks would list it as 2.5-4-4-8. and yeah if you are shooting for a 200x11 setup i would think that would be an ok timing setting. the best think to do though is to check in the area where you set you memory timings and see if it has some presets i.e. - default, by speed, aggresive, turbo, optimized.... something like that. if it does just set it to by speed or optimized or something that is similar. that way your motherboard will set the timings automatically depending on the speed of the cpu. just leave it like that until you get your cpu to overclock to your desired speed and then you can go back and try tightening the timings if you like. and you should be fine with 1.75v and even more.
 
I wouldn't worry too much about CPU getting damaged. It's cheap to replace, under 100 buckx.

Long before you get your CPU damaged, you will experience instabilties along the way. Instability will come much sooner than getting damaged... do you let it run at 100C? 80C? i doubt so.

maybe electromigration is what might be considered.. for some. High Vcore & high temp for a long............ time. but you wouldn't wanna let it run hot. Keep it cool. I keep mine at 11~ 14C@ 2.7~2.8ghz@ above 2v.
 
well....... not everyone wants to shell out a $100 bucks all the time if they can keep from it. i dont let my cpu get over 53C.
 
Back
Top