Are you using the latest Cinebench 2024? If so what test were you running CPU, GPU or both at the same time?

Originally posted by wing0zero:

Are you using the latest Cinebench 2024? If so what test were you running CPU, GPU or both at the same time?

i forgot, but something something(GPU) , and yes cinebench24 from maxson website
or maybe because iam running it on HDD? the fan is so fast like a jet engine made me lost spirit to test again

Well I'm just wondering if both tests at the same time was too much for some part of the PC, I wouldn't suspect RAM as it's not stressed with Cinebench really, maybe it was a one off crash unrelated, you would have to test again to see if it's repeatable.

If it doesn't like a full load maybe a CPU test like Prime95 and FurMark at the same time might do something similar.

If you can stress the system to hard crash every time I would suspect PSU or Motherboard as a first guess, but that's only if it's repeatable you can't go off one event.

Do ram test first... memtest 86. Free works fine. 1 pass is enough. Check PC can shuffle ram. If it fails don't bother to do the others. Ram is faulty check your settings.

Do prime95 15mins. Usually enough to see that CPU can do math.

The furmark or 3d mark. Free is fine. Test graphics.


My overclock PC checker suit always works.

PSU true. It can even be too low wattage

Originally posted by Carlsberg:

Possible the psu is at fault.

Thats what I was thinking.

OP try using hwinfo or other app that can monitor power draw of both the CPU and GPU. Could be the psu being pushed too far for too long.

If it happens again you could also use afterburner to cap the power draw of the gpu. Drop it 5% and see if it fixes it. If it does increase in1% increments till it crashes nothing go back 2%

i would play with the voltage of the gpu. you could void the warrenty. leave it stock. goto your bios and try stock bios reset to "optimized defaults" from there make only necessary changes for successful boot.

Try this. Otherwise its a new power supply

undervolt all you want, it is not dangerous. overvolting however is a different story. thing will not be damaged by having too little power, only run slower. i once had to undervolt my pc, because of a power problem in my house. no damage but, it ran more slowly until the problem was fixed.

undervolting make it less stable and less heat

which can corrupt the os if its crashing
sfc /scannow can fix most of the time

too little power does not try to make it run slower
thats underclocking, by lowering multi, fsb and disabling turbo/boost/pbo

Originally posted by HIVEmind:

i would play with the voltage of the gpu. you could void the warrenty. leave it stock. goto your bios and try stock bios reset to "optimized defaults" from there make only necessary changes for successful boot.

Try this. Otherwise its a new power supply

Why woulkd we undervolt the gpu ? that's going to tell us what ? nothing.

Apparently the machine is running fine, the question was;

What would/could have made it unstable for the cinebench test causing the crash. Realistically,.., that machine would need to be in front of most to determine that with the results drawn from tests.

Unclear if this is a laptop or desktop.

isnt cinebench explicitly testing the CPU? Not sure why thatd hurt the GPU. Any idea what the thermals were looking like during the test?

A hard crash like that certainly sounds like a power thing. If the thermals were in a typical range id say yeah a PSU issue for sure.

If the CPU caused a crash like that without thermal throttling id say check that you have all the connections filled. By the cpu there is a couple 4 pin connectors. if only one of them is populated then whoop. for overclocking it needs the other one conected

how about complete specs?

post a cpuz validation link
http://www.cpuid.com/softwares/cpu-z.html
cpuz -> validate button -> submit button
it will open a browser, copy the url (address) and paste it here

and psu brand/model/age

Here's a rule of thumb when it comes to PCs; "it's stable until it demonstrates that it's not". That's generally how stability is determined because there is no singular fail proof software that will tell you if it's stable.

In practice, most people run a limited amount of software (compared to the total amount that exists), and if it's stable with their normal routine, they presume it is factually stable. For practical purposes, that might be fine, but you need to be open to the idea that if you do encounter something that is unstable, you have to be open to the possibility that an unstable system is the cause, despite that fact "other things work fine" or "it worked fine with the same things before". None of that removes the possibility of an unstable system being the cause. Instead, when most people begin to encounter something that is not stable, the kneejerk reaction is usually "it's the software/game/its engine that's unstable, not my PC". Sometimes that's true, and sometimes it's not.

A PC shutting off in the way you described (screen goes Black, fans go to full) is absolutely an instability, yes. What was the exact source of that instability is hard to say. My guess would be GPU (which includes drivers) or PSU are the most likely source of such a crash.

poor psu can seem stable til specific loads, then its not

some gpus and psu combos are unstable because of the way the gpu loads at specific fps when using vsync, switching between boost and idle clocks at a frequency the psu cant deal with fast enough