Execute a freeze and restart while under CPU and GPU load, avoiding BSOD.
Execute a freeze and restart while under CPU and GPU load, avoiding BSOD.
I searched extensively online but couldn't find sufficient guidance from seasoned users. Recently, I upgraded my CPU and performed stress tests to verify stability. Specifications: CPU i9 11900kf, cooler Noctua NH-D12L, motherboard NZXT Z590 (BIOS P1.50), RAM 2x16GB at 3200MHz, GPU MSI RTX 3060 Ventus with two OC SSDs, PSU Corsair SF750 (750W), case Cooler Master CMP 520L with Noctua fans. I ran Windows 11 and followed Intel updates for drivers and software. Using default settings with XMP and adaptive boost enabled, the Cinebench R23 multicore test and FurMark performed well; temperatures stayed within safe ranges. The system completed without BSOD during testing, though GPU temperatures peaked at 170W and 80.6°C on the hotspot (69°C average). HWinfo indicated PL1 and PL2 limits, but I managed to push CPU power slightly higher before freezing after two minutes of stress. From the data, it appears my PSU might be struggling beyond 125W PL1, despite being under its rated capacity. The PSU fan is operating normally, and airflow from the front case to the unit seems adequate. It seems I’m drawing around 300-400W during peak loads, which is still well within the PSU’s capacity. I suspect a PSU limitation rather than hardware constraints, but a more experienced eye might spot other factors. Thank you for your consideration.
Run the same tests using the PSU hanging from the back rather than inside the case. I've noticed similar thermal problems with the power supply when it's not sealed, and fixing it by letting it vent freely outside resolved the issue.
Sorry for the delayed reply, I only had time to test it today. Thanks for your feedback! I managed to remove the PSU and place it on the table outside the case. The test completed successfully. Appreciate your questions! I’m curious about how sensitive PSUs are to temperature changes. The unit didn’t overheat when placed outside, even though it could “breathe” freely. Is that a significant difference? Also, would using a higher wattage PSU be beneficial? Would extra capacity make a noticeable impact or remain similar? Lastly, I noticed my “Turbo boost power window” was set to 128 seconds—right around the time the system would freeze and restart with the case’s PSU. Is there a link between these events, or is it just coincidence? Edited February 16, 2024 by Bbbbaun
The only case I've encountered of this was a Thermaltake GF3 1600W that consistently overheated in a Fractal Design Torrent. The issue stemmed from low build quality and a subpar case design, which were resolved by allowing free airflow—though I ultimately returned it as defective, as confirmed by Thermaltake support. You might try flipping the power supply to create positive pressure inside your case, ensuring it faces downward to avoid suction from filters. It’s unlikely to interfere with your RTX 3060 fans, but adjusting the bottom and front fans up slightly could help.
I first set the fan up facing upward inside the case. Later I flipped it so it draws in from outside when I suspected a temperature issue. Now it's back with the bottom filter out to improve airflow, and I've lowered the CPU power limits back to default. I might reach out to Corsair for their recommendations. Thanks for your assistance!