PC freezes / crashes when i start cinebench r23 multicore test. It also crashes while exporting in premire pro

[Ezio29]

i7-11700f

3060ti

2 x 16 DDR4 

gigabyte H510M H V2

 

MSI MAG A650BN PSU 650 W

 

I dont really experience high temps or crashing while playing games but whenever i go on to do a stress test or just do an export on premire pro it crashes or freezes immediately.

I have checked the thermal paste and it looks good.

 

Is there an issue with my PSU? Is this a CPU problem?

 

 

[podkall]

18 minutes ago, Ezio29 said:

Is there an issue with my PSU? Is this a CPU problem?

Actually it might be a Motherboard problem, your motherboard is quite cheap but your CPU is 8 core beast.

 

Cinebench R23 pulls all cores at 100% utilization because they're crunching rendering of an image, in games you don't always get 100% CPU utilization of all cores, so games don't crash.

 

https://www.hwinfo.com/download/

 

You can check the motherboard temps to see if the VRMs aren't overheating or getting really warm, alternatively the MB could just not have enough power for CPU.

 

My advice, for now don't worry about it, since it only happens for R23 and Premire.

 

Premire probably utilizes lot of CPU when doing it's thing, some limits in BIOS could help CPU draw less, which could help the motherboard handle big demand, of course this will probably slow down the CPU and performance a little bit.

[Ezio29]

Thanks for the reply, I tried HW info already but the issue is it crashes almost immediately and there is no time to check the temps.

But I'll keep your recommendation in mind.

[Spicyegg_234]

download xtu, but instead of overclocking set a really low current limit (like 50-75 A) to see if it doesn't crash with less current

[Spicyegg_234]

2 minutes ago, Spicyegg_234 said:

41 minutes ago, podkall said:

Actually it might be a Motherboard problem, your motherboard is quite cheap but your CPU is 8 core beast.

 

https://www.hwinfo.com/download/

 

You can check the motherboard temps to see if the VRMs aren't overheating or getting really warm, alternatively the MB could just not have enough power for CPU.

funny quote tool

try pointing a box fan at the motherboard

[191x7]

3 hours ago, Ezio29 said:

i7-11700f

3060ti

2 x 16 DDR4 

gigabyte H510M H V2

 

MSI MAG A650BN PSU 650 W

 

I dont really experience high temps or crashing while playing games but whenever i go on to do a stress test or just do an export on premire pro it crashes or freezes immediately.

I have checked the thermal paste and it looks good.

 

Is there an issue with my PSU? Is this a CPU problem?

 

 

Latest Bios?

Which CPU cooler?

RAM running XMP?

Which case and how many case fans?

[podkall]

9 hours ago, Ezio29 said:

Thanks for the reply, I tried HW info already but the issue is it crashes almost immediately and there is no time to check the temps.

But I'll keep your recommendation in mind.

and if you play games, have you tried leaving HW info running while playing those?

[IseeFractals]

What's the power limit for your cpu? PL 1 is 65w, but the power draw for PL2 can get up to 288w... and unless you manually changed it, MB manufacturers tend to default limits to their highest setting in pursuit of claiming performance. 

Rendering tasks and benchmarks hit the cpu with 100% load, which means a maxed out power draw. It could be that your PSU is underpowered, it could be that your VRMs are overheating or simply can't handle the current, or it could be that the high load is outpacing the limits of your CPU cooler. It could also be unstable memory speed/timings. Hell, i had a similar crashing issue a couple months ago that was solved by updating the firmware on my SSD. 

Also, look at HWmonitor and look for the 12v/5v/3.3v rails. Voltages should be within +/- 0.2v at most, and stable within that range. If voltages are dropping or spiking...that could be the issue.

But given that it's an intel chip, my suspicion is the underlying cause is at least related to the power limit. Boot into your bios, disable XMP, set the PL to 65w . Boot into windows, open HWmonitor and make sure the clock/power sensor details are visible and then run  r23. If it survived, go back into the bios and load up XMP settings. Boot back into windows, HWmonitor and benchmark again. If it survives, then start pushing up the CPU. 

The CPU could have degraded, the motherboard could be degrading, or the memory, or the chipset, or the PSU. All you can really do is drop everything to minimum, and start stepping up settings for individual components until you start seeing crashes again, then dial it back a bit. As my uncle used to say "step through it, figure it out" If you can't survive a benchmark run, it's obviously not stable, and ideally you should be ensuring that it can sustain 100% load for an hour at a minimum without crashing, but as that's a bit of a tall ask, ensure at least a 10 minute r23 run.