Shoutbox

Ok, so my PC has been turning off when performing fairly intensive tasks such as youtube videos, playing games (TF2 mainly) and erm, etc.

So I thought it was an issue with the graphics card overheating. It's a 9500GT and the operating temperature is up to 105 degrees C. I've been monitoring it for quite a while now, and it does get pretty high (I've seen up to 104), but it has still been turning off without hitting the max temp.

I saw a post on here today about running memtest86 (I was already worried it was a bad sector (or whatever) of RAM). So after burning it and running it after a restart, it gets about 37% through the test and the PC turns off.

So:

1) I'm guessing it's safe to assume it is the RAM
2) I have 2 RAM sticks installed; is there any way of narrowing it down?
3) Is there any way to fix it or is buying new RAM the only (or cheaper/easier) way

Allow me to oppose, but it sounds to me more likely your CPU
1) is not being cooled properly (check your heatsink or redo the heat paste) (you can also boot into windows and use temperature monitoring programs to see if your cpu is overheating)
 
2) is not supplied with enough power (maybe you forgot some 2x2 or 2x4 power supply plug)

---

Oh and your questions..
Most motherboards boot with only one memory module installed. Get the manual and look for which DIMM slot the module should sit (should be the first slot though)

Most memory module manufacturers have their product under lifetime warranty (at least so in Taiwan) and you probably can have your modules fixed/exchanged-for-new at no charge. Depends on your memory module manufacturer though.

If you do need to replace the ram I just got my hands on 6 1GB sticks that I ain't using.

quote:

---

Originally posted by Spunky
1) I'm guessing it's safe to assume it is the RAM

---

No, in fact, several indicators are pointing to something else than RAM.

It's unusual for the PC to just shutdown during memtest. If it was bad memory memtest will just report errors, but the PC will almost always keep running.

You say the maximum temperature for your graphics card is 105C (which seems correct according to Google), and you have seen it report up to 104C? That's obviously way too high, and seems to me you are hitting the max temp. You'll probably never see it say 105 because that's the temperature when it will immediately shut down.

Except for shutting down, did you see memtest report any actual errors? If not I would forget the idea of RAM being the issue completely for now and focus on getting your GPU temprature down.

Is it overclocked in any way? Badly applied heatsink? Sure fans are spinning?

Thanks Mnjul! I just checked with SpeedFan and the temp does seem to be pretty high on GPU, Temp1 and Core. Running a stress test it raised to over 100 degrees and shut-off again.

Unfortunately, doing an inspection of the fan (which I noticed was the source of a noise I'd been hearing lately) ended up with it losing a couple of "fins" making it unbalanced. I can get one in the next couple of days for around £5.

I've managed to move the rear fan to the side by a vent and funnel thing so it directs air straight at the CPU and I managed to get a couple of 4 inch usb fans off of a laptop stand and plugged them into the motherboard for now.

I'll let you all know if the new fan(s?) fixes the problem