I have a 3-4 year old computer with a P5NSLI from Asus with an E6600 chip (I believe)
In any case, over the past 2 months, my computer has spontaneously shut down (running windows XP, but it has shut off once during hardware detection of the BIOS). Every time it has done so it has been within 20 minutes of startup, but it sometimes does so repeatedly. If it makes it beyond the 20-30 minute mark, it'll stay up for multiple days if I let it. My system logs are empty for any strange events (it doesn't log an error before shutting off)
Today, on a whim, I downloaded HWMonitor from CPUID along with Speedfan. Both are reporting ultra high temps
(TMPIN0/Temp1: 96-97 C, TMPIN2/Temp3: 89 C)
My GPU is also pretty high (NVIDIA GeForce 8600 GT: 61 C)
I noticed my voltages are wonky. My -12V sometimes has a lower absolute value than my -5V and both are usually very close to each other. At one point in this typing they were both -7.35 and have the same max value (-4.35) and similar min value (-12.10 V for -5 and -11.97 for -12)
Both programs say that I have a first and 3rd fan, but not a second fan. One fan is going about 2400+ RPM, the other around 1000.
Do I have a failed power supply, a failed fan (perhaps because of the PS), or faulty temperature sensors? Is the power off caused by a spike in the temp above 100 C?
Do a visual of all the fans, and check the shutdown temp in the bios or the shutdown setting with the asus monitor program, if you're using it. Some power supplies shut the fan off when it's not needed. The minus settings are less important, and yours may be fine. Some hardware monitor programs don't give accurate readings. I only go by the bios readings. I found coretemp to be more accurate. My only concern is the cpu readings. I also clean the fans with a canned air product that contains special solvents that don't leave a residue. They sell for $4-5 at Frys.
I cracked open the case, and vacuumed and then manually dusted flat surfaces and the fans. The heatsink on the processor was under a pretty thick layer of dust. All my main fans (2 case fans, the power supply fan, the chip fan) were functioning properly, but I was unable to check the video card fan - not a lot of clearance there - which could explain its poor performance. One of my case fans has very little clearance in my desk hutch, but I've been using this general setup for several years now (predating this computer).
According to my power supply, there is no -5V, so I'm guessing it was just double-reporting the -12V supply with a slight reading delay.
I've been up for about 15 minutes on lightish load (running BOINC, so processor at 100%, but no graphics load to speak of. Current temp is 55/37/51 with ACPI@40 and cores at 64/60 and gfx at 62. Remote case temp is registering as 57 F (so...15 C or so )
We'll see how things do after an hour or two of load.