Need Help With SiSoft Sandra... Possible Cooling Problem

[Scroatdog]

I set Sandra to a continuous loop so I can "burn-in" my system. I'm also trying to troubleshoot my PSU voltage.

Sandra stops mid-way through the first run with the error message saying "CPU Cooling Thermal Resistance" 1.#JC/W

What does that mean???? Is my system going to FRY if I keep trying??

My idle temps seem to be 36-37C, with a case temp of 38, and CPU kernal temp of 70.

I forgot to "burn-in" the system when I put it together on Saturday.....

Anyone??? Please HELP

 

[eklass]

CPU kernal temp? huh?

my athlon 2000 with a volcano 9 goes around 43C under full load

not sure about you intel-type people though

 

[MistaTastyCakes]

Not sure what a kernel temp is, but if your idle is that 36-37, your load should be in the 50's at the most, I'd guess. If you wanna stress test your system, ditch Sandra and download Prime95. Burning in a system is a myth, more of a placebo than anything else. Get Prime95 and memtest86 to test if it's stable under load or not, and take it from there. For temp monitoring, use Motherboard Monitor 5 to see your true CPU temps in real time.

Sandra does some things alright, but stress testing is not one of them.

Hope this helps a little

 

[Scroatdog]

I tried the torture test on Prime95, and got this:

Intel(R) Pentium(R) 4 CPU 3.00GHz
CPU speed: 3014.47 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: 8 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 version 22.12, RdtscTiming=1
Best time for 256K FFT length: 7.883 ms.
Best time for 320K FFT length: 10.496 ms.
Best time for 384K FFT length: 12.670 ms.
Best time for 448K FFT length: 15.338 ms.
Best time for 512K FFT length: 17.323 ms.
Best time for 640K FFT length: 22.479 ms.
Best time for 768K FFT length: 27.298 ms.
Best time for 896K FFT length: 33.714 ms.
Best time for 1024K FFT length: 36.383 ms.
Best time for 1280K FFT length: 51.382 ms.
Best time for 1536K FFT length: 64.035 ms.
Best time for 1792K FFT length: 78.373 ms.
[Tue Apr 29 21:54:23 2003]
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
[Tue Apr 29 22:20:27 2003]
FATAL ERROR: Rounding was 0.4990234375, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.

Is my system jacked up or what??? I did a reboot and set my memory back to SPD of 2.5-7-4-4 and got the same thing.

What is WRONG?

 

[OulOat]

Uh, what are the temps that mbm reports? Your CPU is probably overheating, you are lucky it's a P4 instead of an Athlon.

 

[WarCon]

Just for giggles slow your memory down to a lower multiplier than 1:1 or run Memtest.

If you are really hitting CPU core temps of 70C then your CPU is throttling most likely to stay right there and not go higher. Are you using the stock heatsink? Are you sure its install correctly? (I.E. are all four corners latched?) Any additional cooling fans? Is it hot in your room?

Some things to think about.....

P.S. Read somewhere that this board is very picky with memory.

 

[Jarhead7236]

uncheck terminate on overheat/failure box on the burn in wizard...that'll keep it from stopping. intel processors have built in throttling to keep them from getting burned out due to overheating. throttling dials the cpu back to lower the cpu temp...and the system slows down but doesn't burn up...something to that effect..it works...i've read and seen it for myself. you may need to up your dram voltage in the prime95 situation....have you run memtest86 to check your ram? try it and see if there are errors...

 

[Scroatdog]

Originally posted by: Jarhead7236
uncheck terminate on overheat/failure box on the burn in wizard...that'll keep it from stopping. intel processors have built in throttling to keep them from getting burned out due to overheating. throttling dials the cpu back to lower the cpu temp...and the system slows down but doesn't burn up...something to that effect..it works...i've read and seen it for myself. you may need to up your dram voltage in the prime95 situation....have you run memtest86 to check your ram? try it and see if there are errors...

Yeah. Did that and Sandra did her thing. Then I got a warning from MBM5 and SilenTek telling me that my CPU has reached its temp limitss. SilenTek reported it ad 305.6C (yeah right), and MBM5 reported it as -47C.

I think I have a bum power supply and it's making my system unstable. The +12v and the +3.3v lines fluctuate erratically, all the time.

I will try memtest when I get home. I downloaded the one where you can make a bootable CD with. Now I just need to figure out how to make th CD bootable. Never done that before with Nero. I'll keep you posted, but my idle temps have CONSISTENTLY been at 36C, load up to 42, and as high as 51C after gaming. It displays CPU kernal temp at 70C all the time, but I have no idea what the kernel temp is used for. Case temp is always at 36-38C.

Keep in mind, this is a brand new system and the first system I've ever built, so theres a strong possibilty I screwed someting up. I'm pretty sure I got just te right amount of AS3 on the core. Spent 20 miuntes doing it too, making sure it was perfectly even and thin. Also rubbed AS3 into the heatsink as per Artic Silver's instructions.

I'll keep you all posted. Should see my new PSU by Friday (fingers crossed)

 

[OulOat]

Uh, load = 100% CPU usage consisently. So how can your CPU be hotter after gaming?

 

[Scroatdog]

Originally posted by: OulOat
Uh, load = 100% CPU usage consisently. So how can your CPU be hotter after gaming?

Ok, so it's 51. I was typing too quickly and not paying attention.