Hi all,
I believe my fairly new GPU is rapidly giving up the ghost, but would like your opinion on if I should look into getting it repaired/fixed!
It's a 2070 Inno3D. Generally my PC has been fine, running games smoothly etc. A few weeks ago I could hear what might be described as coil whine when playing games.
As of yesterday however, games have been stuttering to the point they are unplayable! E.g. GTA, Mordhau, which before were playing fine. FPS drops to 20 or less constantly.
Browsing and general video watching seems OK, as does playing games that are not graphically demanding.
Other weird issue is that textures seem to be taking longer to load, with the typical fuzzy textures everywhere. Temperature is in the 40-50 deg range generally.
I ran Unigine Heaven benchmark which reported, avg FPS 159.9, score 4028, Min FPS 16.7. Not as bad stuttering, but there were definite split second freezes throughout.
Now the main thing that points to an issue is I ran memtestG80, and it is reporting large error numbers, whereas for a friend of mine it is consistently 0 errors.
My example runs:
50 iterations over 128 MiB of GPU memory: 45331 errors
50 iterations over 128 MiB of GPU memory: 76333 errors
50 iterations over 500 MiB of GPU memory: 345649 errors
100 iterations over 500 MiB of GPU memory: 593244 errors
100 iterations over 500 MiB of GPU memory: 528422 errors
The program is not more specific where the errors are, as in the breakdown it says 0 errors on each iteration, but the total errors are accumulating - screenshot:
https://imgur.com/a/Oad0Urp
I would say the latter is a very strong argument that the GPU is buggered, but thought you guys might have some insights?
Thanks!
All components purchased around Nov. '18 from OC UK, so probably all still under warranty (better be!).
Specs:
AMD Ryzen Threadripper 1920X
Asus Prime X399-A mobo
Inno3D GeForce RTX 2070 Twin X2
Corsair Vengeance RGB Pro 32GB total
Corsair 750w Gold PSU
Enermax AIO cooler
No overclocking.
Other side-observations:
The CPU seems quite hot, too hot to touch on the cooler sides close to the mobo, and sensors showing 60 idle, 70ish under load, but then I read that the 1920X series reports temps that are in fact 27 deg lower than true junction temp! So potential issue, but unlikely.
https://www.guru3d.com/articles-pages/amd-ryzen-threadripper-1920x-review,8.html
Another minor issue is that yesterday after a power-cycle the mobo got stuck booting, code 68 'PCI Host Bridge initialization'. But that only happened once and not repeatable; possibly a one-off but thought I might mention it.
I believe my fairly new GPU is rapidly giving up the ghost, but would like your opinion on if I should look into getting it repaired/fixed!
It's a 2070 Inno3D. Generally my PC has been fine, running games smoothly etc. A few weeks ago I could hear what might be described as coil whine when playing games.
As of yesterday however, games have been stuttering to the point they are unplayable! E.g. GTA, Mordhau, which before were playing fine. FPS drops to 20 or less constantly.
Browsing and general video watching seems OK, as does playing games that are not graphically demanding.
Other weird issue is that textures seem to be taking longer to load, with the typical fuzzy textures everywhere. Temperature is in the 40-50 deg range generally.
I ran Unigine Heaven benchmark which reported, avg FPS 159.9, score 4028, Min FPS 16.7. Not as bad stuttering, but there were definite split second freezes throughout.
Now the main thing that points to an issue is I ran memtestG80, and it is reporting large error numbers, whereas for a friend of mine it is consistently 0 errors.
My example runs:
50 iterations over 128 MiB of GPU memory: 45331 errors
50 iterations over 128 MiB of GPU memory: 76333 errors
50 iterations over 500 MiB of GPU memory: 345649 errors
100 iterations over 500 MiB of GPU memory: 593244 errors
100 iterations over 500 MiB of GPU memory: 528422 errors
The program is not more specific where the errors are, as in the breakdown it says 0 errors on each iteration, but the total errors are accumulating - screenshot:
https://imgur.com/a/Oad0Urp
I would say the latter is a very strong argument that the GPU is buggered, but thought you guys might have some insights?
Thanks!
All components purchased around Nov. '18 from OC UK, so probably all still under warranty (better be!).
Specs:
AMD Ryzen Threadripper 1920X
Asus Prime X399-A mobo
Inno3D GeForce RTX 2070 Twin X2
Corsair Vengeance RGB Pro 32GB total
Corsair 750w Gold PSU
Enermax AIO cooler
No overclocking.
Other side-observations:
The CPU seems quite hot, too hot to touch on the cooler sides close to the mobo, and sensors showing 60 idle, 70ish under load, but then I read that the 1920X series reports temps that are in fact 27 deg lower than true junction temp! So potential issue, but unlikely.
https://www.guru3d.com/articles-pages/amd-ryzen-threadripper-1920x-review,8.html
Another minor issue is that yesterday after a power-cycle the mobo got stuck booting, code 68 'PCI Host Bridge initialization'. But that only happened once and not repeatable; possibly a one-off but thought I might mention it.