|
Have you checked the RAM? You could, e.g., try running Prime95 "Blend" test for a few hours. If you get an error (e.g. a rounding error) you may have a bad stick of RAM. Memtest86+ also tests RAM, but from what I have read Prime95 stresses the memory in ways Memtest86+ does not. Have you tried taking out your video card and using your onboard video for a few days? If there are no errors at least you've narrowed down the problem some. If onboard video is stable, you could try running your card in the other graphics PCIE x16 slot. If your system is stable, your primary PCIE slot may be bad. If the problem reappears, perhaps that's enough of a reason to RMA the card? Obviously if you have a spare PSU you can swap in, or another system you can test the card in, you have additional options to narrow down the problem.
|
#
¿
May 6, 2016 01:59
|
|
|
|
| # ¿ May 22, 2024 00:10 |
|
|
Ularg posted:You're talking about this, right? Yes. That's old info but the details seem mostly right. You don't need to run multiple instances manually, the newer versions of Prime95 use multiple threads to test all your cores. Keep an eye on temps (core temps primarily) periodically as the test runs; things will get pretty hot and this can help diagnose an overheating problem. I use HWMonitor for this--I like how it keeps track of min and max values.
|
|
#
¿
May 6, 2016 22:02
|
|
|
Those are good idle temps. Your processor will slow itself down if it gets into the range where damage is likely, so your risk of permanently harming the CPU is low. If the system gets into "borderline too high" territory, you may want to keep running the test for a bit longer and see if that's what's causing your instability. If, on the other hand, you're getting into "clearly far too high" territory, I'd stop Prime95 (your temps should go back down immediately) and probably look into reseating your heatsink. On temps, I'm not an expert and you might want to Google for yourself, but I would say that max temps in the 70s are probably OK, over 80 is too high, 90+ is where your processor will start throttling itself to cool down. If it were my computer I'd stop the test if the cores started getting in the 85 range, you shouldn't see anything that high if you're not overclocking. One important thing to consider is that Prime95 will probably work your processor harder than anything you're likely to encounter in normal usage, so slightly high temperatures aren't that big a deal as long as your system is stable. Good luck!
|
|
#
¿
May 7, 2016 01:09
|
|
|
You can check the log in the directory where you have Prime95, the log is called results.txt. Does it indicate any errors during your run?
|
|
#
¿
May 7, 2016 22:05
|
|
|
Hm, looks like you did not throw any errors, so unfortunately the test hasn't narrowed down your problem. I'd say it's worth using the onboard video for a while and see if your issue persists, that should help tell you if it's the card that's the problem.
|
|
#
¿
May 7, 2016 22:42
|
|
|
|
| # ¿ May 22, 2024 00:10 |
|
|
From your description, the lockups aren't correlated with the video card being stressed (it was presumably idle while you were running Prime95 and you still froze), so you can probably just use your system the way you do normally. Sure, run some games on low settings, run the Cinebench benchmark (the graphics part) if you feel like it.
|
|
#
¿
May 8, 2016 04:45
|
|