|
Problem description: My system will lock up for a few seconds at a time at seemingly random. Video halts, the mouse won't move, and audio goes completely silent. This seems to happen roughly once every hour, regardless of what activity I'm performing, and I can tell when it's about to happen, because the sound stutters for a split second, then a few seconds later the freeze occurs. Nothing shows up in the Event Viewer to give a clue, but I've left task manager open and noticed that my CPU, GPU, and all my hard drives (both primary and secondary) saw a very brief performance spike when the freeze occurs. The memory sees no change. EDIT: Since making this post, the spikes in Task Manager aren't consistent. I just had a freeze with no performance spikes shown. Attempted fixes: I've tried updating every possible driver I can think of, (Chipset, SATA, Audio, Graphics, etc), Run windows updates, run checks on all my drives, run memtest86, and Reinstalled Windows via the reset option that allows me to keep personal files. I've even tried playing with some BIOS settings, such as forcing my M.2 as PCIe instead of "Auto" for example, as well as just resetting them to all default. Recent changes: Prior to the current issue I'm posting about, I was dealing with my PC randomly restarting for no apparent reason. I built this current system at the end of 2017, and the restart issue popped up at the end of last year. I kind of just lived with it for a while due to laziness and ignorance as to how to solve it. For the record, Event Viewer showed Error ID 41, but my PSU voltages seemed fine. As part of my trouble shooting I reset the BIOS by removing and reseating the battery, and had to remove my video card in order to gain access to the battery. It didn't solve the issue, so next I updated my BIOS firmware, and allowed the built in BIOS setup wizard to tweak my settings. It completely stopped the reboots from happening, but left me with the freezing problem. -- Operating system: Windows 10 64-bit System specs: CPU: Intel i7-8700 Motherboard: ASUS Z370-P Graphics card: NVidia Geforce GTX 1080 Memory: CORSAIR Vengeance LPX 16GB (2 x 8GB) 288-Pin DDR4 SDRAM DDR4 3000 Power Supply: EVGA SuperNOVA 650 G3, 80+ GOLD, 650W Fully Modular Primary Hard drive: SAMSUNG 960 EVO M.2 500GB NVMe PCI-Express 3.0 x4 Secondary Hard drives: 2 x Western Digital SATA drives, 1 x Western Digital SSD, 1 Seagate External USB Storage Drive that's always connected. Dual monitor setup: Older model LG using DVI connection, plus a newer Dell with DisplayPort connection. Other Peripherals: Avermedia LiveGamer Portable 2+ capture card (Connected to a PS4), Presonus AudioBox (USB Device that allows me to use a XLR microphone) Location: United States I have Googled and read the FAQ: Yes. Free Gratis fucked around with this message at 21:47 on Jul 3, 2020 |
#
?
Jul 3, 2020 20:20
|
|
|
|
| # ? Apr 26, 2024 16:21 |
|
|
By disconnecting some of the peripherals you can isolate the issue more. The capture card, the audiobox and the GPU are the three things I'd try removing initially (or one at a time).
|
|
#
?
Jul 4, 2020 23:30
|
|
|
Zogo posted:By disconnecting some of the peripherals you can isolate the issue more. I did some further troubleshooting before seeing this post and may have solved the issue. I managed to catch the "system interrupts" process spike in the Task Manager during one of the freezes, and installed Latencymon to see if I could pinpoint the source of it. Latencymon's meters pegged out during a freeze and the biggest offender seemed to be ACPI.sys. Googling didn't help too much because all posts I could find seemed to be related to laptops, as the driver governs the battery's power management. Dxkrnl.sys, Wdf01000.sys, and nvlddmkm.sys, also showed high latency. I then went into the BIOS and changed a CPU power setting to Best Performance, and my system went over 2 hours without a freeze. I went to bed with Latencymon running all night, and none of the meters pegged out which tells me a freeze didn't occur. They did get about half full, but I'm not savvy enough to know if small periods of higher latency is just normal. I went into the NVIDIA control panel and changed a power setting in there to "Best Performance" as well to see if it further alleviates the issue. I'm going to continue monitoring for the day to see if I get a freeze, and if so then I'll be following your advice, which I greatly appreciate. This has been one of the weirdest and most frustrating issues I've ever dealt with.
|
|
#
?
Jul 5, 2020 13:46
|
|
|
I would update your drivers, use the following app to get them. https://www.microsoft.com/en-us/p/snappy-driver-installer/9p1mxhqwkkt2 ACPI usually doesn't cause issues like that unless there is something going on with hardware (unlikely as its rather new) or there are driver issues. Worst case, you get the latest drivers for your computer. Let us know if that helps out.
|
|
#
?
Jul 6, 2020 18:00
|
|
|
DeepBlue posted:I would update your drivers, use the following app to get them. I haven't had another freeze since I changed that one BIOS setting, but I went ahead and took a look at this program, and sure enough, it said several drivers were out of date, and Latencymon seems to show more stability. I appreciate the help.
|
|
#
?
Jul 8, 2020 22:07
|
|
|
Free Gratis posted:I haven't had another freeze since I changed that one BIOS setting, but I went ahead and took a look at this program, and sure enough, it said several drivers were out of date, and Latencymon seems to show more stability. I spoke too soon. Last night LatencyMon caught ACPI acting up again when I was away from my machine, and this morning I got a freeze while I had no monitoring up. I rolled back to a restore point before I used SnappyDriverInstaller and am back to no freezes.
|
|
#
?
Jul 9, 2020 19:29
|
|
|
|
| # ? Apr 26, 2024 16:21 |
|
|
Free Gratis posted:Primary Hard drive: SAMSUNG 960 EVO M.2 500GB NVMe PCI-Express 3.0 x4 One of the rarer, but possible, causes of your issue is if the SSD garbage collection algorithms fail. I've seen this more than once - the SSD is otherwise healthy, but doesn't do automated background garbage collection and TRIM functions during idle time. I don't know what causes it - hardware, firmware, no idea. But when it happens, it will manifest in exactly this way. The underlying reason is that the SSD reaches the end of its writable blocks, then the entire system freezes in place while the SSD performs an on the spot garbage collection routine to open up new blocks. This can happen even if the drive is only partially full because wear-leveling algorithms make sure the entire storage space of the drive is written to evenly. Example: Samsung Magician shows my drive has a total of 2 TB written so far. It's a 500 GB drive which means so far the entire area of the drive has been written to 4 times over, even though the drive is 70% free space. So if I had the problem I described above, the next time my system writes hit the last usable block, the system would freeze waiting on the AHCI controller, while the drive freed up a few more blocks. Then the system would be perfectly fine for a time, until the next time the drive needs empty blocks. I have a few SSDs I've set aside with this problem, which are OK for unimportant storage and otherwise stable, but make for a lousy OS drive. I've had good luck with the 860 EVO SATA drives, you've got a faster one but any SSD can have this problem. Have you checked the SSD firmware revision?
|
|
#
?
Jul 13, 2020 23:02
|
|
