|
Problem description: In early 2019 I built my current system. As part of assembling it, I researched and set up a mild overclock on my i9 9900k to enable 5GHz all-core clock. The system appeared to work fine, including under maximum load, and while the temperatures while running ASUS Realbench in all tests mode got into the 90s celsius, I never hit those temps in normal use so I wasn't worried. Sometime last year I started contributing to World Community Grid, which if I forgot to pause it while playing certain graphics-intensive games would cause a BSOD reliably enough that I eventually disabled WCG on it. Which brings me to this week. I ordered another SSD and some more RAM (exact same SKU as the RAM I already installed) and put everything in. Because I have an NH-D14, any upgrades require me to take the cooler off, so that means clean and re paste every single time, so I did that. However, after installing the components and reassembling everything, I can't even get through a 10-minute all-core compile without a BSOD. Argus Monitor shows temperatures hitting right around 86 celsius when the BSOD occurs. I re-did the thermal paste again, making absolute sure I was doing it properly, but zero effect on the temps or crashes. I also cleaned out dust when I had the case open, but there wasn't an extreme amount built up either in the filters or on the components. My next move is going to be to back off the OC, but I'm posting here to see what else I could possibly look in to track down the issues. Not sure it matters, but the BSODs give me IRQL_NOT_LESS_OR_EQUAL or PAGE_FAULT_IN_NONPAGED_AREA randomly, I'm sure that's just a quirk of the actual instability, rather than related. I'm fairly certain it's not the new RAM, because I've been able to take the system up past 48GB used without issues, and the memory usage while running this compile test is below 20GB. Attempted fixes: As above, repasted the CPU again to make sure it was done right, and will be backing off the overclock and trying again. Will also try a side-off run just to see if there's any effect on temps. Recent changes: Installed RAM and an additional NVMe SSD, which requires a full CPU cooler removal/reinstall to get at the components. No other configuration changes. -- Operating system: 64-bit latest consumer build of Windows 10 Professional. System specs: Board: ASUS ROG STRIX Z390-E CPU: Intel i9 9900k (small OC to 5GHz, I believe 1.3V but I have to check) Cooler: Noctua NH-D14 Case: Corsair Carbide 275R (I know the airflow isn't great since it's not an open front but haven't had problems before.) GPU: EVGA GeForce 1070 FTW3 Memory: Corsair CMK32GX4M2B3200C16W Vengeance LPX 32GB DDR4 3200 C16 (I haven't even touched RAM overclocking, all stock) Power Supply: EVGA Supernova 750 G2, 80+ Gold 750W NVMe Drives: 1TB Samsung 970 Pro (original), 4TB WD SN750 (new), other SATA drives I won't list unless you think it's relevant. Location: What country are you in? US I have Googled and read the FAQ: Yes
|
#
?
Jun 6, 2021 22:10
|
|
|
|
| # ? Apr 29, 2024 06:10 |
|
|
You could try running memtest overnight at some point to check RAM health: http://memtest.org/ If that shows as okay you should make sure you're on the latest motherboard BIOS. PS Probably not an HD issue but it never hurts to run CDI to check HD health: https://osdn.net/projects/crystaldiskinfo/downloads/75303/CrystalDiskInfo8_12_2.exe/
|
|
#
?
Jun 13, 2021 05:38
|
|
