Crash every few hours "Gpu instance has been suspended..."

[Happy BOB]

Crash window https://imgur.com/a/ea8WvLs
4080s with 19-14900k here. The game runs actually pretty smoothly on my ultra-wide screen after I installed Lyall mod . However, after a few hours play. "gpu instance has been suspended..." will pop and the game closes after I clicked ok. I re-run the game, everything is fine for like 2-3 hours and "gpu instance..." pops again. It pops at pretty random timing. I would say most of them are in the middle o f boss fight cut-scene and its the only type of message I am getting. Just want to know whether other people are having the same pop window and is there other specific way to solve it? (Try those all: everything up-to-date/file integrity check/delete shade file/increase virtual Ram etc.)

[iameatingjam]

Hmm. I've had that both when I undervolted my 4090 too much, and when I left the program running in the background too long (I'm on a 14700Kf, and just got a new one today)

Is your gpu undervolted? I hope so because the other likely cause is cpu degradation. Try turning xmp off and reducing all your cores my 300mhz. If that solves the problem, then its cpu degradation. ( that includes the single and dual core boosts too, don't use those).

You on the latest bios? One with 0x12B ucode?

[s0t3lg]

could be because of your mod or denuvo, i play on a 3090 and its fine for the most part however when i first load the game for the day i usually have funny things happening with my cpu and gpu heat and resources used. after i exit out and wait a few hours and load back in it stops. this also doesnt happen when i put my computer into sleep mode and then wake it up the next day.

[Happy BOB]

Originally posted by iameatingjam:

Hmm. I've had that both when I undervolted my 4090 too much, and when I left the program running in the background too long (I'm on a 14700Kf, and just got a new one today)

Is your gpu undervolted? I hope so because the other likely cause is cpu degradation. Try turning xmp off and reducing all your cores my 300mhz. If that solves the problem, then its cpu degradation. ( that includes the single and dual core boosts too, don't use those).

You on the latest bios? One with 0x12B ucode?

Yeap Bios is up-to -date. Version 1666 (11.08 MB 2024/09/30) for Z790 GAMING WIFI7. And I did do the degradation test including Intel® Processor Diagnostic Tool/cinebenchmark. The test pass and temp/volt/hz/score all look fine, which I think at least it has not been affected by degradation that much. That is why I am confused. I might try to see whether its the setting of that Lyall mod since I am using the latest version of that, which is different than the one in those ff16 ultra-wide mod tutor videos.

[Smokey]

Its probably the mod causing this. You can however try to see if using Gpu scaling helps with this.

Nvidia control panel > Adjust desktop size and position. And change "Perform scaling on" from Display to GPU.

[Happy BOB]

Originally posted by Smokey:

Its probably the mod causing this. You can however try to see if using Gpu scaling helps with this.

Nvidia control panel > Adjust desktop size and position. And change "Perform scaling on" from Display to GPU.

I will try it when I get time playing again

[daprile410]

Same here, it started couple weeks back. it's happened maybe 3-4 times. i'm on NG, completed dlcs all side quests and on final boss before NG+ so roughly 4 crashes in ~idk 50+ hours of gameplay.

4070Super
Ryzen 7 7700
5700mhz 32 gig ram

i don't run dynamic resolution and run balance/performance with no visual issues to really speak of. I'm running the other graphics settings at max on a 1440p display usually setting FPS cap at either 72 or 120. I don't have anything overclocked and all other games run fine.

Going to try and turn scaling back on. it was off due to horrid visual artifacts with Elden Ring foliage. If it helps I'll report back. Seeing as it's crashed very little in a large play time sample it might not happen for another 10+ hrs of gameplay. TBH, the previous driver update or two back never caused this.

[McGeorgeOfTheJungle]

Originally posted by s0t3lg:

could be because of your mod or denuvo, i play on a 3090 and its fine for the most part however when i first load the game for the day i usually have funny things happening with my cpu and gpu heat and resources used. after i exit out and wait a few hours and load back in it stops. this also doesnt happen when i put my computer into sleep mode and then wake it up the next day.

Nothing to do with Denuvo, the engine is quite demanding and can get to 80C+ quite quickly on most high end GPUs and that is with of course DLSS etc. But it is more likely his mod yes.

[Suminod]

I Get it on a 4090 FE and 7950X3D so its not specific to anything. No OC or anything just normal and it happens usually after an hour or two.

[MrSoul]

Originally posted by Suminod:

I Get it on a 4090 FE and 7950X3D so its not specific to anything. No OC or anything just normal and it happens usually after an hour or two.

driver issue maybe? Only say this after helping a friend troubleshoot SM2 and his 7900 card, game kept borking on him while trying to play coop till we tried out some beta drivers. Very similar error message he'd get, somethingsomething gpu.

Wonder rolling back, or looking at any beta releases would help?

[Happy BOB]

Originally posted by daprile410:

Same here, it started couple weeks back. it's happened maybe 3-4 times. i'm on NG, completed dlcs all side quests and on final boss before NG+ so roughly 4 crashes in ~idk 50+ hours of gameplay.

4070Super
Ryzen 7 7700
5700mhz 32 gig ram

i don't run dynamic resolution and run balance/performance with no visual issues to really speak of. I'm running the other graphics settings at max on a 1440p display usually setting FPS cap at either 72 or 120. I don't have anything overclocked and all other games run fine.

Going to try and turn scaling back on. it was off due to horrid visual artifacts with Elden Ring foliage. If it helps I'll report back. Seeing as it's crashed very little in a large play time sample it might not happen for another 10+ hrs of gameplay. TBH, the previous driver update or two back never caused this.

I think I am close to the game’s ending now and I haven’t found the solutions yet. It happens once 2-3hours, obviously more frequently, to me but I can accept that since I won’t come back to it after I finish the story.

[iameatingjam]

Originally posted by Happy BOB:

Originally posted by iameatingjam:

Hmm. I've had that both when I undervolted my 4090 too much, and when I left the program running in the background too long (I'm on a 14700Kf, and just got a new one today)

Is your gpu undervolted? I hope so because the other likely cause is cpu degradation. Try turning xmp off and reducing all your cores my 300mhz. If that solves the problem, then its cpu degradation. ( that includes the single and dual core boosts too, don't use those).

You on the latest bios? One with 0x12B ucode?

Yeap Bios is up-to -date. Version 1666 (11.08 MB 2024/09/30) for Z790 GAMING WIFI7. And I did do the degradation test including Intel® Processor Diagnostic Tool

Your processor has to be in really bad shape to not pass that test. There is no test for the 13/14 gen degradation issue. Especially ones that have only minor-moderate degradation.

But if your processor is degraded, downclocking by 300mhz all cores and turning off xmp is highly likely to stabilize it. So if you want to find out if this is this is the issue, do that. If you want false confidence, run the intel intel processor diagnostic tool test.

Same with cinebench, its not really relevant. Thats never a workload that raptor lake has struggled with. It struggles with decompression primarily. Though symptoms do vary wildly.

[Happy BOB]

Originally posted by iameatingjam:

Originally posted by Happy BOB:

Yeap Bios is up-to -date. Version 1666 (11.08 MB 2024/09/30) for Z790 GAMING WIFI7. And I did do the degradation test including Intel® Processor Diagnostic Tool

Your processor has to be in really bad shape to not pass that test. There is no test for the 13/14 gen degradation issue. Especially ones that have only minor-moderate degradation.

But if your processor is degraded, downclocking by 300mhz all cores and turning off xmp is highly likely to stabilize it. So if you want to find out if this is this is the issue, do that. If you want false confidence, run the intel intel processor diagnostic tool test.

That is why I did a lot other tests like cinebench (R23 R15 2024), cpuz and use hwmonitor to check all stats. They do look pretty good and stable during idle or stress status. I wish this were a cpu degrade issue so that I can just replace it without question. I am coming to the end of this game anyway so I just hope the post can give help to other people. (‘-‘)7

[iameatingjam]

Originally posted by Happy BOB:

Originally posted by iameatingjam:

Your processor has to be in really bad shape to not pass that test. There is no test for the 13/14 gen degradation issue. Especially ones that have only minor-moderate degradation.

But if your processor is degraded, downclocking by 300mhz all cores and turning off xmp is highly likely to stabilize it. So if you want to find out if this is this is the issue, do that. If you want false confidence, run the intel intel processor diagnostic tool test.

That is why I did a lot other tests like cinebench (R23 R15 2024), cpuz and use hwmonitor to check all stats. They do look pretty good and stable during idle or stress status. I wish this were a cpu degrade issue so that I can just replace it without question. I am coming to the end of this game anyway so I just hope the post can give help to other people. (‘-‘)7

Guess you missed my edit where I mentioned that cinebench is nearly as irrelevant as the intel tool as raptor lake has never had a problem with with cinebench, it moreso has a problem with decompression and bursty workloads (like gaming).

Cinebench isn't going to break until things are past the point of no return (or you're running an unsteady overclock/undervolt or something). But okay if you really don't want to do the one thing you can do to rule out this problem that should take less than 5 minutes to set up... then fine I guess. You do you.

[Celcius]

It's crashing for me every few hours too. 10700k + 3090 with no mods.