Is my GPU failing/ should I stop folding? Video output issues

[Tropical Snowman]

Hello there, in the last couple months I've been experiencing seemingly random black screens that could be resolved by switching ports on the GPU (it has 3 Displayports).
A web search said it could be a sign of a failing GPU, but also driver issues, PSU issues, cable issues... lots of things. Haven't seen any artifacts, so I didn't give it much thought, undervolted my GPU to stress it less and moved on.

Now, more issues happened:
- my desktop wallpaper disappeared (turned black) - fixed it with a reboot
- my desktop icons turned bigger and their background turned black when hovered with mouse - fixed with a reboot
- my desktop icons turned bigger again with black background - this time the reboot only fixed the background, icons remained large

I'm on Ryzen 7 7700, 32gb DDR5 ram, AMD RX 6700. Windows 11 IoT LTSC. Been folding on this GPU since March.

I'll be trying to rule out software issues by updating to latest update available for Windows 11, i.e.: 24H2 x64 (KB5070773) (26100.6901).
Then, in case it'll persist, I'll uninstall GPU drivers with DDU and reinstall them.

I'm worried I could damage my GPU (or accelerate its failure) but I don't want to stop Folding unless necessary.

Any advice?

PS: I just seem to have solved another issue, don't know if that's related: viewtopic.php?t=43277

[muziqaz]

Your system and windows is broken.
Reinstall windows first. Then see what else is breaking

[Tropical Snowman]

By reinstall you mean a clean reinstall - formatting and installing windows anew? It's my daily pc so I'd avoid it unless necessary

[muziqaz]

By reinstall you mean a clean reinstall - formatting and installing windows anew? It's my daily pc so I'd avoid it unless necessary

Yes, because it seems your install has been corrupted to the point of no return
The combination of your issues is pointing to broken OS. If your system continues to behave similar with fresh OS, then your hardware is dying

[toTOW]

It could be failing system RAM too ... did you run a few Memtest86+ passes to check it ?

[Tropical Snowman]

No, I didn't. I'll do now.

Do you suggest other tests?

[muziqaz]

Y-cruncher

[Tropical Snowman]

Memtest86 failed 1 of 48 tests, with 2 errors. Is my RAM bad? And what do you suggest doing now?
I purchased all the components of this PC last March...

Here's detail:

Result summary
Test Start Time 2025-10-22 12:48:01
Elapsed Time 3:08:31
Memory Range Tested 0x0 - 10000000000 (1048576MB)
CPU Selection Mode Parallel (All CPUs)
CPU Temperature Min/Max/Ave 46C/61C/53C
Lowest memory speed 6000 MT/s (30-36-36-76)
Highest memory speed 6000 MT/s (30-36-36-76)
# Tests Completed 48/48 (100%)
# Tests Passed 47/48 (97%)
Lowest Error Address 0x1D1FD8668 (7455MB)
Highest Error Address 0x1D1FD8670 (7455MB)
Bits in Error Mask 0000000000800000
Bits in Error 1
Max Contiguous Errors 1
CPUs that detected memory errors { 10 }
Bits in Error
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x . . . . . . . . . . . . . . . . . . . . . . .
63 0
Test # Tests Passed Errors
Test 0 [Address test, walking ones, 1 CPU] 4/4 (100%) 0
Test 1 [Address test, own address, 1 CPU] 4/4 (100%) 0
Test 2 [Address test, own address] 4/4 (100%) 0
Test 3 [Moving inversions, ones & zeroes] 4/4 (100%) 0
Test 4 [Moving inversions, 8-bit pattern] 4/4 (100%) 0
Test 5 [Moving inversions, random pattern] 3/4 (75%) 2
Test 6 [Block move, 64-byte blocks] 4/4 (100%) 0
Test 7 [Moving inversions, 32-bit pattern] 4/4 (100%) 0
Test 8 [Random number sequence] 4/4 (100%) 0
Test 9 [Modulo 20, random pattern] 4/4 (100%) 0
Test 10 [Bit fade test, 2 patterns, 1 CPU] 4/4 (100%) 0
Test 13 [Hammer test] 4/4 (100%) 0
Last 10 Errors
2025-10-22 14:15:03 - [Data Error] Test: 5, CPU: 10, Address: 1D1FD8668, Expected: 5EDDCCDA, Actual: 5E5DCCDA
2025-10-22 14:15:03 - [Data Error] Test: 5, CPU: 10, Address: 1D1FD8670, Expected: 5EDDCCDA, Actual: 5E5DCCDA

Here's the rest of the info, if it might be useful to you:
System Information
EFI Specifications 2.90
System
Manufacturer Gigabyte Technology Co., Ltd.
Product Name B650M AORUS ELITE AX ICE
Version Default string-WCP
Serial Number Default string
BIOS
Vendor American Megatrends International, LLC.
Version F37
Release Date 07/31/2025
Baseboard
Manufacturer Gigabyte Technology Co., Ltd.
Product Name B650M AORUS ELITE AX ICE
Version x.x
Serial Number Default string
CPU Type AMD Ryzen 7 7700 8-Core
CPU Clock 3794 MHz [Turbo: 5341.1 MHz]
# Logical Processors 16 (8 enabled for testing)
L1 Cache 16 x 64K (304247 MB/s)
L2 Cache 16 x 1024K (152054 MB/s)
L3 Cache 1 x 32768K (123347 MB/s)
Memory 31993M (60542 MB/s)
RAM Configuration DDR5 6000MT/s / x2 Channel / 30-36-36-76 / 1.100V
Number of RAM SPDs detected 2
SPD #0 16GB DDR5 EXPO 1Rx8 PC5-48000
Vendor Part Info Team Group Inc. / UD5-6000 / 010440D3
JEDEC Profile 4800MT/s 40-40-40-77 1.1V
XMP Profile 1 6000MT/s 30-36-36-76 1.35V
XMP Profile 2 5600MT/s 40-40-40-84 1.2V
EXPO Profile 1 6000MT/s 30-36-36-76 1.35V
SPD #1 16GB DDR5 EXPO 1Rx8 PC5-48000
Vendor Part Info Team Group Inc. / UD5-6000 / 010440CD
JEDEC Profile 4800MT/s 40-40-40-77 1.1V
XMP Profile 1 6000MT/s 30-36-36-76 1.35V
XMP Profile 2 5600MT/s 40-40-40-84 1.2V
EXPO Profile 1 6000MT/s 30-36-36-76 1.35V
Number of RAM slots 4
Number of RAM modules 2
DIMM A1 Empty slot
DIMM A2 16GB DDR5 1R PC5-48000
Vendor Part Info TeamGroup / UD5-6000 / 010440D3
SMBIOS Profile 6000MT/s 1.1V
DIMM B1 Empty slot
DIMM B2 16GB DDR5 1R PC5-48000
Vendor Part Info TeamGroup / UD5-6000 / 010440CD
SMBIOS Profile 6000MT/s 1.1V

[muziqaz]

Double check if RAM voltage is actually 1.35V instead of 1.1V stated in few places in your post. Once you set it to 1.35V manually, observe if the voltage really holds up, even in BIOS you should be seeing the voltage. If it is shows a bit lower than 1.35V, add the difference, and save. Then run Memtest again

[Tropical Snowman]

I first tested if it was an Expo issue, set Expo 2 instead of Expo 1 and Memtest this time passed. Guess RAM is fine, but unstable on Expo 1?

Should I run Y-cruncher now?

[muziqaz]

I don't see EXPO 2 profile in your comment. What does it have different from EXPO 1?

[Tropical Snowman]

If EXPO Profile 1 is 6000MT/s 30-36-36-76 1.35V; Expo 2 is 5600 MT/s CL36-38-38 (if I recall correctly). They're the two standard RAM overclock profiles, usually you'd use EXPO 1, but mine apparently is unstable, or there's some voltage problem with the pre-made profile, as you saw. Instead of tweaking it manually, I just selected the other pre-made profile on my motherboard, which is a bit lighter and more stable (I guess).

Also just tried OCCT, a 15 mins GPU Variable test and a 15 mins VRAM test showed no errors.

Maybe it was all related to the RAM instability? I can do some more test, just to be sure.

[BobWilliams757]

Stability is key for folding, but often a couple memory errors won't cause problems either.

I had some memory with major issues at one point, and never got any video or folding instability at all. But if your video display issues go away, you might have found an answer. But I would suspect problems with the GPU first.

[muziqaz]

If EXPO Profile 1 is 6000MT/s 30-36-36-76 1.35V; Expo 2 is 5600 MT/s CL36-38-38 (if I recall correctly). They're the two standard RAM overclock profiles, usually you'd use EXPO 1, but mine apparently is unstable, or there's some voltage problem with the pre-made profile, as you saw. Instead of tweaking it manually, I just selected the other pre-made profile on my motherboard, which is a bit lighter and more stable (I guess).

Also just tried OCCT, a 15 mins GPU Variable test and a 15 mins VRAM test showed no errors.

Maybe it was all related to the RAM instability? I can do some more test, just to be sure.

Run through several hours of y-cruncher stress test on that xmp2 profile. If it passes, run memtest86+ overnight to confirm the stability

[Tropical Snowman]

Thanks. I'll run Y-cruncher, but I'm more worried if my GPU has some underlying issue that could make it fail with Folding. Any way to test that?