[SOLVED] Requesting help diagnosing the actual cause of issue

JancariusSeiryujinn

Distinguished
Jul 22, 2011
20
0
18,510
0
I have a pretty much brand new PC build, I9-11900k, 3090, 128gb of RAM. Motherboard is a MSI Z-95 Godlike.

http://speccy.piriform.com/results/pEPqbW7JbUfExDhALJ4FZXI

For the first month or so I had it, I had zero issues. Now, however, I am getting a periodic crash that seems to be associated with my drivers.

View: https://imgur.com/a/WwqLJkw



Attached in the image of some of the crash messages I have gotten. Some games crash without any error message, seemingly more common in older games (example of this is Anno 1404 and 2070 - both games have crashed on this PC without any error output). Newer games usually include a "submit a crash report" dialogue but many don't output any specifics to me.

I have RMA'd the 3090 after the test which Nvidia Support directed me to run showed hardware problems. I've gone through several theories, none of which seemed to have panned out as the issue is still occurring.

Theory one: The issue is related to display port cables, particularly ones over a certain length (past 6 feet). The problem began the day I brought in a replacement monitor. I am trying to have this PC mounted in a rack near my desk, but the cable distance is probably about 12 feet when it is racked. I originally tried a Thunderbolt-hub to the monitor via optical and this actually seemed to work without a problem - except that only one monitor can be linked in the daisy chain. So I had returned the optical TB cable and hub and switched to DP cables - initially 15 feet to have some slack. During initial occurence, mean failure time was 20 minutes from launch of a game to a crash.

I have since tried a pair of DP cables connected via repeaters, and a entirely separate single HDMI cable monitor. Mean time to crash increased significantly but the crashes still occured.

Caveat on the single HDMI cable test - In this test, I had nearly 2 and a half hours of game time prior to crash. I had begun to consider it a stable state and a confirmation that the issue was tied to my long DP cables, so I began shifting things around to plug short DP cables in for that test. The moment I plugged a DP cable into the back, it triggered the crash, even without the other end connected to the monitor. I will run a repeat stability test without plugging anything else in while waiting for replies.

Update: Time to fail this time was only 22 minutes.

Theory two: Software issue - When I brought in the new monitors, I also updated to the latest NVIDIA driver. When the issue began occurring I rolled back the driver, but the issue persisted. With NVIDIA support, I did a full driver clean install of an older driver, issue still occured. Returned to latest driver version, issue still occured. If it is software based, my only other guess is a conflict between ASUS MyCrate software trying to play with the colors on my Mobo or the MSI Center I installed related to my motherboard. I pulled the ASUS stuff out just now while writing this, as I don't need to customize what color lights <Mod Edit> is glowing (I cna't even see it inside the server rack anyway).

Update: Uninstalled all the ASUS stuff and all the MSI parts that interact with the GPU that I could note. Played Star Citizen for nearly 4 hours without a crash, when a crash did happen it could easily have been something buggy happening rather than the card. My stability test ran while connected to only a single HDMI monitor for 8 hours uninterrupted. However, I experienced a crash within minutes when I booted FFXIV.

Theory 3: Internal hardware issue. There could be a hardware issue not in the GPU (or maybe just still in the GPU) that I'm not detecting or aware of and am getting red herring'd away from.

I'm looking for suggestions to help me lock down what is happening on this PC so that I can get back to enjoying myself for the holidays.

Update: Ran https://rtech.support/books/how-to-and-guides/page/dism-and-sfc. DISM found nothing, SFC found some corrupt files and repaired them, but the issue is still occuring.
 
Last edited by a moderator:
Dec 30, 2021
77
8
35
7
The 60hz monitor is just somethign I plugged in for diagnosis and testing - It's an older HDMI monitor I had sitting around so I set it directly on the PC rather than dealing with any potential complications from the long-range DP. My main monitor is a 1440p 144hz, and I have a 1080 144hz secondary and once I get this all sorted out, a 1080 60hz I'm going to put in Portrait mode just for browsing during games.

And yep, updated the BIOS a few days before the issue started.
Ah I see, I thought surely he's not gaming at 60hz with such a monster system haha.

Looking at the game Icarus have you tried running that game in Direct x 12? Add -dx12 to launch options and test. With Final Fantasy there is tons of people having issues with high end GPUS and fatal direct x errors.

Are you having any blue screens of death? any at all in reliability history or any dump files within C:/Windows/MiniDump? You can open these with tools such as WinDbg and BlueScreen View 1.55.
 
Dec 30, 2021
77
8
35
7
Are you running the latest bios? Also bit off topic but for such a beast of a system why only 60Hz monitor?

Reading into the Final Fantasy error a lot of people are saying underclocking their stock overclocked GPUS fixed it. There seems to be a lot of people having issues with Final Fantasy and certain GPUS.
 

JancariusSeiryujinn

Distinguished
Jul 22, 2011
20
0
18,510
0
Are you running the latest bios? Also bit off topic but for such a beast of a system why only 60Hz monitor?

Reading into the Final Fantasy error a lot of people are saying underclocking their stock overclocked GPUS fixed it. There seems to be a lot of people having issues with Final Fantasy and certain GPUS.
The 60hz monitor is just somethign I plugged in for diagnosis and testing - It's an older HDMI monitor I had sitting around so I set it directly on the PC rather than dealing with any potential complications from the long-range DP. My main monitor is a 1440p 144hz, and I have a 1080 144hz secondary and once I get this all sorted out, a 1080 60hz I'm going to put in Portrait mode just for browsing during games.

And yep, updated the BIOS a few days before the issue started.
 
Dec 30, 2021
77
8
35
7
The 60hz monitor is just somethign I plugged in for diagnosis and testing - It's an older HDMI monitor I had sitting around so I set it directly on the PC rather than dealing with any potential complications from the long-range DP. My main monitor is a 1440p 144hz, and I have a 1080 144hz secondary and once I get this all sorted out, a 1080 60hz I'm going to put in Portrait mode just for browsing during games.

And yep, updated the BIOS a few days before the issue started.
Ah I see, I thought surely he's not gaming at 60hz with such a monster system haha.

Looking at the game Icarus have you tried running that game in Direct x 12? Add -dx12 to launch options and test. With Final Fantasy there is tons of people having issues with high end GPUS and fatal direct x errors.

Are you having any blue screens of death? any at all in reliability history or any dump files within C:/Windows/MiniDump? You can open these with tools such as WinDbg and BlueScreen View 1.55.
 

ASK THE COMMUNITY