PC restarts using Cinebench, Premiere Pro and Spider-Man remastered version
PC restarts using Cinebench, Premiere Pro and Spider-Man remastered version
Hello everyone, I previously shared a similar situation but didn’t want to bring up an old post that I thought was resolved. Thank you ahead of time for your patience—it’s going to be a lengthy and detailed update. Please let me know if anything needs more explanation!
The challenge I’ve faced is using the same build (with the specs listed) for a while without problems. Recently, I began editing in Premiere Pro and my PC would restart whenever I rendered or played back video. This reminded me of an earlier problem where starting Cinebench caused the machine to shut down immediately, without any BSOD or error codes.
I fixed the Cinebench issue by updating my chipset drivers. But when I encountered the Premiere Pro problem, it returned. I also noticed similar issues with Spider-Man Remastered—I can’t play the game without it. After the first cutscene, I’d reboot right away, no BSOD. I’ve looked into Event Viewer each time, but the only critical event was a Kernel-Power task (ID 41). All other games and load tests except this one never triggered the problem.
At first, I ran a curve optimizer with negative 5 on all cores, boosted CPU by +100MHz, and enabled PBO limits. I thought these settings might be the cause. So I reset my BIOS to defaults. I also monitored temperatures, which seemed normal for Zen 3 under load.
My setup includes:
- Motherboard: MSI B450M
- CPU: Ryzen 9 5900x
- Cooler: NZXT Kraken X62 AIO
- GPU: MSI RTX 3090 Ventus 3X 24GB
- RAM: Kingston Fury Renegade DDR4 3600MHz CL16 32GB
- PSU: Corsair RM 850W (initially) then upgraded to RM1000x
- Storage: WD S3750 1TB NVMe SSD with heatsink case
- OS: Windows (ATX mATX)
I’ve tried many configurations and settings, but the issue persists in some scenarios. I’m sharing what worked and what didn’t, hoping this helps clarify things.
What seems to help (but doesn’t fix) includes:
- Default BIOS settings
- PBO disabled
- No XMP profile on RAM (2400MHz default)
In a few cases, I managed to run Cinebench fully and even play Spider-Man Remastered, though crashes still occurred later. It feels like either my BIOS or drivers might be the culprit. I’m frustrated but open to suggestions.
Thanks for your time and understanding!
I recommend resetting the BIOS back to its default settings, gradually lowering your voltage below the standard level, then reducing your core clock speeds to their minimum. Perform a test and if it fails to reboot, repeat the steps while slowly increasing the clock and voltage until you locate your target. For initial testing, use Full Cinebench r23 and run a 2-minute session for subsequent checks. You don’t need to execute the full benchmark every time. In the final verification, conduct at least two complete tests and spend a full hour or more running OCCT. These procedures tend to be simpler and more direct with Intel processors, though they should yield comparable results to yours. It’s possible your chip is defective; if you manage to identify the red line, staying below it will resolve the issue. If you can’t detect that line, the problem likely lies further upstream. It’s worth noting this issue could also stem from your motherboard not delivering power correctly to the processor. Someone with a similar setup might be able to offer more tailored assistance.
You might also want to verify if your cooler is properly aligned with the processor. It could be a contributing factor.
Thanks for your input! Could you clarify which setting you're talking about in the BIOS? It seems to be related to the CPU core voltage. I'm still getting familiar with these options, but it sounds like the curve optimizer works by lowering the voltage during load. You mentioned the default idle voltage stays around 1.4v, which makes me wonder if voltage changes are contributing to the problem. That's just a guess, but it could be connected to motherboard or CPU compatibility. You mentioned this is the second 5900x you received with similar issues, so it might be about BIOS compatibility. I'll look into that and let you know what I find. Thanks again!
For stock speeds and Ryzen processors, 1.4v is excessive and could cause overheating during load. This concern comes from testing various Ryzen models, including my 3950x, which can run at or slightly above stock speeds with minimal voltage (around 1.27v) without issues. Some variants handle lower voltages smoothly as well. Reducing voltage through undervolting generally poses little risk if done moderately, but excessive lowering can lead to instability. I recommend adjusting the BIOS settings to stock speed and a voltage close to 1.27v. Avoid relying on monitoring tools for real-time data; instead, observe temperatures during actual workloads. Adjustments should be incremental—small changes rather than large jumps—to pinpoint the right setting. If temperatures rise noticeably, inspect thermal paste and ensure proper application. Even with higher-end chips like the 5900x, sticking to recommended voltages helps maintain stability.
Thanks for the response @Beerzerker! I've completed several tests and it's a bit puzzling. I think I've tried various configurations—BIOS defaults, PBO settings, and even without curve optimizer or boost overclock. It seems there might have been an unnoticed setting that wasn't cleared, or perhaps I missed something during testing.
In my experience, when I set PBO to Advanced in BIOS and disabled the curve optimizer or boost overclock, I still experienced reboots. Even with a boost of 50MHz and without curve optimizer, the system would restart. Negative values for the curve optimizer caused reboots, while enabling boost overclock without optimization also led to restarts.
On the other hand, PBO enabled with Advanced and no curve optimizer resulted in stability. I tried adjusting the CPU voltage manually at 1.27V and without setting a specific speed, but the CPU still ran below stock speeds (around 3.4-3.5 GHz). I'm curious if forcing the CPU to stock speed would prevent this.
I also want to know why Advanced PBO settings seem to trigger reboots. Could it be related to a sudden voltage drop from 1.4V to around 1.1V during operation? I'm not very comfortable with undervolting or overclocking, so I apologize if this is unclear.
Would it be possible to run a stable curve optimizer with custom limits for PPT/EDC? I've tried that too, but it hasn't worked consistently.
Also, do you think certain motherboards or Ryzen 5000 models might behave differently? Should BIOS defaults be adjusted if they're significantly higher than normal? Thanks again for your guidance!
Great to hear! I’ll review the video carefully and check the best settings for your build. Your curiosity about PBO Advanced causing reboots is interesting—though it seems similar to PBO Enabled, there might be subtle differences in how they interact with your system. Keep an eye out for any patterns when everything runs on auto.