System halts entirely when processing increases in demand
System halts entirely when processing increases in demand
Hello. I'm here to share details about a very odd freezing problem that has appeared since switching to an HRR display on my Arc B580. It closely matches what was reported a few days back. The issue is consistent across all checks—no errors in the event viewer or similar tools. The main difference is that I'm using an i5-14400F processor, which is supposed to be unaffected by Intel's known problems. Some people suggest it might be related to power issues, but I'm not there yet. Here are my system details:
- Motherboard: MBO
- CPU: Intel Core i5-14400F
- Graphics: Intel Arc B580 (ONIX AIB model)
- RAM: 2x16 DDR5-5600 from Crucial MBO
- SSD: Asus Prime B760M-A-AX
- HDD: WD Blue 256GB m.2
- PSU: EVGA 450BR
- Under load (mainly gaming), the system occasionally freezes completely.
- Monitors freeze on the latest frame; USB devices briefly dim and then return.
- Speakers sometimes produce pops or cracks.
- After upgrading to a high-refresh monitor, I initially thought it was graphics driver failure, but event viewer showed no issues.
Some have suggested power problems, though I'm running Furmark and Prime95 simultaneously while testing stability. I plan to keep those programs running for a while tonight (CPU locked at PL2) just in case, but I’m not sure if power is the cause. Memory concerns are less likely since I don’t experience BSODs from them. I suspect oxidation might be involved, but the 14400F model isn’t officially affected.
I’ve updated the BIOS to the latest microcode and disabled ASPM. The unusual behavior includes a high virtual memory usage during gameplay (up to 97% on Final Fantasy XV Splitgate 2). Other games I’ve tried show similar spikes, though I don’t play them often. I have various diagnostic tools installed—HWinfo64, Prime95, Cinebench, Furmark, ThrottleStop—and everything seems normal except for the occasional memory strain.
I’ll keep sharing updates as new information comes in.
The best way to prevent the Intel CPU from overheating is by slowing down the P cores by 300-500MHz compared to the stock speed. My 14900K was stabilized for a few months by reducing the 6GHz cores to 5.7, but I recently began experiencing problems again and lowered all P cores to 5.6 GHz, which resolved the issue. Based on my experience, these CPUs are likely to fail if not managed properly.
If it's a new CPU that's installed with a BIOS updated within days or weeks, it's probably fine for the life of the PC. But if you've been running that CPU for months before updating the BIOS, chances are damage to the CPU die has already occured, sadly.
I've experienced two 14900K units, and all of them only ran the patched microcode before shutting down. It seems either Intel or the motherboard is at fault. The CPU attempted to operate at 330W+ even though it should have used the default 253W power plan. It completely disregarded the power limits I set, even when a 253W cap was manually applied, causing it to push 330-350W.
It's a locked chip, and I'm not sure I can adjust the clock settings. The system operates near 125 watts out of its maximum 148 watts (Intel's TDP) if it runs at full load continuously. Do you think this sudden shutdown is due to overheating? I recall experiencing a BSOD a few weeks ago where the CPU stopped responding abruptly, but usually it would show a warning before stopping completely. EDIT: Would running Prime95 all night help determine if the issue is with the CPU?
If manual adjustment isn't possible, reduce the BCLK from 100Mhz to 96-95MHz. The core frequency equals BCLK multiplied by the multiplier per core. A default of 14900K corresponds to 100MHz × 60 = 6GHz. For a 14400F clock, lowering the frequency below 47MHz works by using 95 × 47 ≈ 4.465GHz. You might consider dropping another 100-200MHz for extra safety. Alternatively, use Intel Extreme Tuning Utility, but if it's locked, you're stuck. Disabling hyperthreading and/or E cores could also yield results.
I’m willing to give it a shot. My recent run of prime95 lasted around nine hours and everything worked fine. EDIT: Should I just test this by staying idle or self-botting in games? EDIT: I adjusted the clock multiplier to 92% in ThrottleStop, but my performance dropped from 120 FPS to 30. Could be I’m doing it incorrectly.
Your system experienced frequent crashes and BSODs linked to memory usage. It seems the issue might be related to background data transfers, especially from cloud services or large file operations. I wonder if applications like OneDrive, Google Drive, or other software are handling significant data movement during gameplay or backups. Also, high-quality settings in apps such as Nvidia could be affecting recording performance.
I possess a Steam recording, though the file size isn’t huge. I also use OneDrive occasionally, but it doesn’t get used much, so unusual behavior shouldn’t occur. My machine never restarts by itself without my input. The only clear sign of trouble was one BSOD shortly after launching Windows with a CPU watchdog warning. It seems the CPU simply gave up at that moment. I haven’t experienced this again. Recently, I left Final Fantasy XV idle for several hours and didn’t encounter any issues in other titles, leading me to suspect a PC-related problem—such as memory failure, CPU instability, driver crashes, or graphics issues—that might trigger a crash when the system can’t proceed. Benchmarks, stress tests, and performance checks haven’t caused it either. So far, I’ve only noticed this kind of abrupt shutdown in Final Fantasy XV (both with DXVK for Intel Arc and native DX11) and Splitgate 2. I’m planning to add a list of games and programs that have crashed on this system to the original post.
It occurred once more consecutively within short time, less than an hour after playing FFXV again. It remained stable for days prior until then.