I recently bought an upgrade to my pc which was AMD64 x2 4400, 2x1GB OCZ Dual Channel PC3200 XTC Edition and A8N32 SLI Deluxe mobo
Everything is running smoothly but have a wierd problem
Well, let me start from begining
I put everything in and installed win and all the drivers, etc and installed a speedfan to monitor my temps. It was a bit high up around 45-47 for cpu on idle, so I remembered before with some mobos I had a problem where a bios flash would help, so I did an upgrade of bios to the newest one 1303 dos bios version from ASUS official website. After doing that my temp sensors started showing a decent temp which could be expected from x2 averaging 33 idle and up to 50 on stress, e.g. playing battlefield 2 and such.
Well, after all that was done, I was quite happy my x2 was running cool on Zalman CNPS9500 LED so I decided to give it a bit of oc. The most I could get out of it on 1.450 volts was 2.6 that would boot into windows, but I was not satisfied with it, I knew this baby could do more on this mobo and with this ram.
So, I saught a help from some of you guys here on Anandtech and a guy recommended to me that I should gradually crack into the core buy slowly rasiing the speed and doing some prime time along with it for a period of time until processor gets used to this mobo and all that. So I tried undevolting it and going for 2.5 ghz but it would only give me no errors in prime if I left it at 1.425 which is default volts for this cpu. Anything under and prime would give me error like this:
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
That was with 2 prime instances opened, one for each core, so I left it at default volts and all other settiings which gave me initial 2.2 ghz so I figured I do soome test with prime at default speeds and clock.
I ran two instances of prime, one for each core and I monitored it for a half an hour and no errors appeared. I was a bit happier. Saying that, I went to bad leaving prime running overnight. I did a torture test on both cores with small FFTs settings as Zebo suggested.
So I woke up in the morning I turned on my monitor and saw that my computer froze completely, nothing was working, it was on but win had just froze everything, mouse, ketboard, nothing was working so I did a restart.
Looked at the prime logs and there were no error and such. The tests ran for 5 hours and the computer froze on both cores on the last thing from log:
Self-test 14K passed!
[Sun Sep 17 09:57:15 2006]
Self-test 16K passed!
Self-test 14K passed!
[Sun Sep 17 09:54:33 2006]
Self-test 16K passed!
It did not report any errors and yet, my pc was frozen.
Anywho, that was not he only thing that bothered me. After a restart I noticed something else wrong. My idle temps have gone up and now it is iddling at 41-44 celsius instead of nice and cool 32-34.
To mention, temperature during prime was about 56-58 while running two primes for each core.
So folks, that would be my problem of the day.
Anyone have any suggestions, insights? I would greatly appreciate any input towards this problem.
Thanks in advance
Everything is running smoothly but have a wierd problem
Well, let me start from begining
I put everything in and installed win and all the drivers, etc and installed a speedfan to monitor my temps. It was a bit high up around 45-47 for cpu on idle, so I remembered before with some mobos I had a problem where a bios flash would help, so I did an upgrade of bios to the newest one 1303 dos bios version from ASUS official website. After doing that my temp sensors started showing a decent temp which could be expected from x2 averaging 33 idle and up to 50 on stress, e.g. playing battlefield 2 and such.
Well, after all that was done, I was quite happy my x2 was running cool on Zalman CNPS9500 LED so I decided to give it a bit of oc. The most I could get out of it on 1.450 volts was 2.6 that would boot into windows, but I was not satisfied with it, I knew this baby could do more on this mobo and with this ram.
So, I saught a help from some of you guys here on Anandtech and a guy recommended to me that I should gradually crack into the core buy slowly rasiing the speed and doing some prime time along with it for a period of time until processor gets used to this mobo and all that. So I tried undevolting it and going for 2.5 ghz but it would only give me no errors in prime if I left it at 1.425 which is default volts for this cpu. Anything under and prime would give me error like this:
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
That was with 2 prime instances opened, one for each core, so I left it at default volts and all other settiings which gave me initial 2.2 ghz so I figured I do soome test with prime at default speeds and clock.
I ran two instances of prime, one for each core and I monitored it for a half an hour and no errors appeared. I was a bit happier. Saying that, I went to bad leaving prime running overnight. I did a torture test on both cores with small FFTs settings as Zebo suggested.
So I woke up in the morning I turned on my monitor and saw that my computer froze completely, nothing was working, it was on but win had just froze everything, mouse, ketboard, nothing was working so I did a restart.
Looked at the prime logs and there were no error and such. The tests ran for 5 hours and the computer froze on both cores on the last thing from log:
Self-test 14K passed!
[Sun Sep 17 09:57:15 2006]
Self-test 16K passed!
Self-test 14K passed!
[Sun Sep 17 09:54:33 2006]
Self-test 16K passed!
It did not report any errors and yet, my pc was frozen.
Anywho, that was not he only thing that bothered me. After a restart I noticed something else wrong. My idle temps have gone up and now it is iddling at 41-44 celsius instead of nice and cool 32-34.
To mention, temperature during prime was about 56-58 while running two primes for each core.
So folks, that would be my problem of the day.
Anyone have any suggestions, insights? I would greatly appreciate any input towards this problem.
Thanks in advance