Re: Premiere crashes if "Hardware-accelerated GPU sched." enabled

System Specs:

Amd Threadripper 3960x 24 core

RTX 2080ti connected to 2 4k 60hrz displays

256GB Ram 

SSDs all the way down, some attached USB storage

Asrock STRX40 Motherboard

I have left this setting off for years and tried to turn it back on today, just to see if my system was still buggy with this feature turned on. I've learned a few things since then. Its pretty bad, since I believe this feature is default ON with windows 10 installs. I see the OP also has a 2080ti, I wonder if this card specifically has some sort of bug that just was never addressed. 

Without fail within a few min of using premiere I get a black screen and a GPU driver crash. This time it was just relinking a lot of media while archiving a project. 

If someone from adobe could respond I can send more logfiles etc. @Kevin-Monahan do you know who might be a good person to tag in this discussion?

Is there something buggy about my specific hardware config or has anyone else left this setting on recently and not run into these issues? I guess it doesn't hurt to leave off but I feel like After Effects runs about 10% better with it left on. 

When I look into the Premiere Pro logfiles I get a massive amount of "CUDA_ERROR_ILLEGAL_ADDRESS" messages when the GPU Driver crashes.

Ticks = 469438 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": false, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuEventQuery\", \"deviceMemory\": {\"outstandingSize\": 540928, \"allocatedSize\": 747577600, \"pooledCount\": 6, \"allocatedCount\": 15, \"softPoolLimit\": 5905350656, \"hardPoolLimit\": 8858025984}, \"hostMemory\": {\"outstandingSize\": 16589344, \"allocatedSize\": 33721120, \"pooledCount\": 8, \"allocatedCount\": 10, \"softPoolLimit\": 1073741824, \"hardPoolLimit\": 2147483648}, \"videoMemory\": {\"budget\": 10757341184, \"usage\": 4897423360}}"
Ticks = 469454 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": true, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuEventRecord\", \"deviceMemory\": {\"outstandingSize\": 17129728, \"allocatedSize\": 747577600, \"pooledCount\": 9, \"allocatedCount\": 15, \"softPoolLimit\": 5905350656, \"hardPoolLimit\": 8858025984}, \"hostMemory\": {\"outstandingSize\": 16589344, \"allocatedSize\": 33721120, \"pooledCount\": 7, \"allocatedCount\": 10, \"softPoolLimit\": 1073741824, \"hardPoolLimit\": 2147483648}, \"videoMemory\": {\"budget\": 10757341184, \"usage\": 4897423360}}"
Ticks = 469454 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": false, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuMemcpyHtoDAsync\", \"deviceMemory\": {\"outstandingSize\": 17129728, \"allocatedSize\": 747577600, \"pooledCount\": 9, \"allocatedCount\": 15, \"softPoolLimit\": 5905350656, \"hardPoolLimit\": 8858025984}, \"hostMemory\": {\"outstandingSize\": 16589344, \"allocatedSize\": 33721120, \"pooledCount\": 8, \"allocatedCount\": 10, \"softPoolLimit\": 1073741824, \"hardPoolLimit\": 2147483648}, \"videoMemory\": {\"budget\": 10757341184, \"usage\": 4897423360}}"
Ticks = 469454 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": false, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuMemFree\", \"deviceMemory\": {\"outstandingSize\": 0, \"allocatedSize\": 747577600, \"pooledCount\": 10, \"allocatedCount\": 15, \"softPoolLimit\": 5905350656, \"hardPoolLimit\": 8858025984}, \"hostMemory\": {\"outstandingSize\": 16589344, \"allocatedSize\": 33721120, \"pooledCount\": 8, \"allocatedCount\": 10, \"softPoolLimit\": 1073741824, \"hardPoolLimit\": 2147483648}, \"videoMemory\": {\"budget\": 10757341184, \"usage\": 4893294592}}"
Ticks = 469454 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": false, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuMemFree\", \"deviceMemory\": {\"outstandingSize\": 0, \"allocatedSize\": 747577600, \"pooledCount\": 15, \"allocatedCount\": 15, \"softPoolLimit\": 5905350656, \"hardPoolLimit\": 8858025984}, \"hostMemory\": {\"outstandingSize\": 16589344, \"allocatedSize\": 33721120, \"pooledCount\": 8, \"allocatedCount\": 10, \"softPoolLimit\": 1073741824, \"hardPoolLimit\": 2147483648}, \"videoMemory\": {\"budget\": 10757341184, \"usage\": 4889231360}}"
Ticks = 469454 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": false, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuArrayDestroy\", \"deviceMemory\": {\"outstandingSize\": 0, \"allocatedSize\": 714400000, \"pooledCount\": 14, \"allocatedCount\": 14, \"softPoolLimit\": 5905350656, \"hardPoolLimit\": 8858025984}, \"hostMemory\": {\"outstandingSize\": 16589344, \"allocatedSize\": 33721120, \"pooledCount\": 8, \"allocatedCount\": 10, \"softPoolLimit\": 1073741824, \"hardPoolLimit\": 2147483648}, \"videoMemory\": {\"budget\": 10757341184, \"usage\": 4889231360}}"
Ticks = 469454 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": false, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuArrayDestroy\", \"deviceMemory\": {\"outstandingSize\": 0, \"allocatedSize\": 714383616, \"pooledCount\": 13, \"allocatedCount\": 13, \"softPoolLimit\": 5905350656, \"hardPoolLimit\": 8858025984}, \"hostMemory\": {\"outstandingSize\": 16589344, \"allocatedSize\": 33721120, \"pooledCount\": 8, \"allocatedCount\": 10, \"softPoolLimit\": 1073741824, \"hardPoolLimit\": 2147483648}, \"videoMemory\": {\"budget\": 10757341184, \"usage\": 4889231360}}"
Ticks = 469454 <37256> <GF.CUDAError> <3> CUDA Error, result: "CUDA_ERROR_ILLEGAL_ADDRESS", detail: "{\"first\": false, \"error\": \"CUDA_ERROR_ILLEGAL_ADDRESS\", \"detail\" : \"cuMemFree\", \"deviceMemory\": {\"outstandingSize\": 0, \"allocatedSize\": 713859328, \"pooledCount\": 12, \"allocatedCount\": 12, \"softPoolLimit\": 5