I have been getting random reboots which seem to point to a hardware issue. I've tried running Prime95 to hopefully diagnose the problem. My machine is stable and runs forever on the "small fft" test which is mainly a cpu test as I understand it. However, my machine reboots fairly quickly on the "in-place large fft" test. I have tested my ram in all variations with MemTest86 as well as the Windows Memory Diagnostic and it tests out fine after numerous passes and many, many hours. What does this scenario point at as the hardware issue? Here are my specs (btw, my cpu runs about 58 C at idle and Prime95 gets it up to 71 C or so):

Operating System : Windows XP Prof SP2
CPU/Processor : Athlon xp 2600+
Motherboard : gigabyte 7nf-rz (f3 bios)
Hard Disk : maxtor, Space: 120 GB
Video Card : Nvidia GeForce FX 5200
Video RAM : 128 MB
Sound Card : Soundblaster Live! Value

With an idle temp of 58C I would say it`s likely an overheating problem.

Check that your system is dust free.

Check your cooling fans.

Remove the heatsink, and clean off the old thermal pad/paste with a little isopropyl alcohol.

Apply a very thin layer of new thermal paste, and reattach the heatsink.

See HERE for instructions on how to apply thermal paste.

See if that helps.

Regards Howard

Athlons don't like temps over 55c for one thing. I would seriously check and reapply your thermal paste. Prime95 can usually be stabilized with a bump in vcore but at your temps, I wouldn't recommend it until the temp problem is solved.

Thanks guys. I just recently applied thermal paste to get my temp down to the 58C range on idle. I was running about 10 degrees hotter. Is there any way to tell if I've irreparably damaged my cpu? By the way, my latest run at the small fft Prime 95 test pooped out after about 9 hours with a reboot.

Somethings deffinitely not right there.

No way should your cpu idle at 58 degrees celcius.

Excuse the question. You did put the heatsink on the right way round didn`t you?

You only applied a small amount of thermal paste?

Is the cpu fan working correctly?

Download the free Everest programme from HERE

See what the temp sensor says.

Regards Howard

The heatsink is on correctly. It may not have been seated well the first time around and I just used the thermal pad that originally came with the heatsink originally. When I re-seated it though, I applied the thermal paste to the die in a very thin layer after carefully cleaning the pad off the die and the heatsink. I'm pretty sure I did everything as well as possible. Is the stock heatsink/fan on this processor really cheap maybe? I'm not even sure this is causing my random reboots, but I've ruled out just about everything else so I'm working on temps now to see if it helps. Incidentally, I used a credit card to remove the pad from the cpu die...that is okay isn't it?

Anyway, here is what Everest reports (Is it true that Gigabyte boards are notorious for overstating temps...perhaps these aren't accurate?):

Sensor Properties:
Sensor Type ITE IT8712F (ISA 290h)

Temperatures:
Motherboard 31 °C (88 °F)
CPU 58 °C (136 °F)
Aux 39 °C (102 °F)
Maxtor 6L120P0 31 °C (88 °F)

Cooling Fans:
CPU 3444 RPM
Chassis 2250 RPM

Voltage Values:
CPU Core 1.71 V
+2.5 V 2.69 V
+3.3 V 3.20 V
+12 V 12.29 V
+5 V Standby 3.09 V
VBAT Battery 3.09 V

I can`t think why your temps would be so high, unless of course your mobo`s sensor is way off.

Zip 5 or 6 of your latest minidumps together, and attach them here.

You can find them in C:/windows/minidump.

Regards Howard

All my minidumps are 0x7f. I notice that your profile lists you as using an Athlon XP 2600+. Are you using the stock heatsink/fan? What are your temps? Thanks for all your help!

All your minidumps crash with an unknown image. They all reference prime95, and all have a bugcheck of 7F.

0x0000007F: UNEXPECTED_KERNEL_MODE_TRAP

One of three types of problems occurred in kernel-mode: (1) Hardware failures. (2) Software problems. (3) A bound trap (i.e., a condition that the kernel is not allowed to have or intercept). Hardware failures are the most common cause (many dozen KB articles exist for this error referencing specific hardware failures) and, of these, memory hardware failures are the most common.

Yes I`m using the stock amd heatsink, and fan. My idle temps are around 28c and under full load I get around 45c to 50c.

I agree that you`ve got some kind of hardware problem. Unfortunately the minidumps don`t say which hardware it is.

Regards Howard

Thanks for the info. How can you tell that the minidump references Prime95? I may as well learn some skills out of all this mess. I have debugged a small bit, but am not very proficient. Your temps are so much lower. I may get a new heatsink and see if it helps.

In the windbg command line type !thread

This will show you what was running at the time of the crash.

Regards Howard

Either temps (most likely) or prime95 just doesnt like you and the hardware. happens all the time, prime just doesnt work on some systems

Does that mean that prime95 fails because it just doesn't work with some configurations even though there is no hardware problem?