I, too, have the GA-MA78GM-S2H, and the AMD Athlon 64 X2 5000+, and I have the 4 GB "Upgrade" Memory kit from OCZ, which is 2 chips @ 2 GB each, DDR2 800. The case is a Thermaltake Lanbox, which is fairly compact, and maybe a consideration in all this. Also, the small board size means that two-thirds of the HSF sits directly under the power supply, with about an 1.5 inches between the top of the HSF to the bottom of the PS. Maybe that effect is not good for cooling (I dunno?!) HSF is a CoolerMaster Vortex 752, and is running at 2368 rpm, fixed. I have disabled both the CPU Smart Fan Control, and AMD Cool & Quiet functionality in the BIOS, per the recommendations at the "
http://www.overclock.net/amd-general/307316-offical-overclocking-guide-amd-socket-am2.html" that I've been studying, but otherwise, I am not overclocking at all (not yet... not with the heat problems I'm having!)
At idle, the temps are as such (per Hardware Monitor):
MB:
TMPIN0 38°C TMPIN1 47°C TMPIN2 92°C
CPU: core #0 34°C core #1 30°C
GPU: 56°C
The other monitoring tools report pretty much the same temps (CoreTemp, Everest, & AMD Overdrive). Idle temps have been stable since I first built the machine and installed Vista Home Premium.
But in the last month or so, and particularly in the last 2 days, almost anytime the CPU gets under heavy load, the temps on the CPU & TMPIN's rises super high, and super fast, and I end up with a thermal shutdown. (Admittedly better than a totally fried system!!)
Since March, I have done many hours of intensive video encoding, with the CPU under heavy load and performing very satisfactorily.
The overheating perhaps has something to do with my upgrading the BIOS to F4 or even F3, but that might be a coincidence.
Any suggestions? I am at the point where I think I need to reseat the HSF, to make sure the thermal paste is well applied, but I think I might need a bigger better fan -- or could this problem be in the NB and/or SB chips overheating and not the CPU? I suspect the CPU myself, because last time I heard the overheat warning from the system, I had time to launch Hardware Monitor and saw the CPU temp reported as high as 100 C, and the TMPIN2 temp made it up to 125 C or something insane like that?
Which raises another question -- I had heard that those TMPIN0 & TMPIN1 were actually the CPU temps, and TMPIN2 was the actual MB temp -- which seems preposterous, or just guilty of an incredibly inaccurate monitoring system.
When I killed the program that was cranking on the CPU, the temps came down, pretty quickly at that.
Any help greatly appreciated. I'm not new to computers, but am a medium-novice when it comes to the deep down hardware issues.
Thank you,
Will in Palm Springs
(and yes, the computer is in an air-conditioned room, grin)