Error in NL construction

aoliva

My simulation end up with this error after running smoothly for more than 4 hours.

*** Exception (JCellDivGpuSingle::Divide)
Text: Error in NL construction. (CUDA error: an illegal memory access was encountered).

Do you have any idea about how to solve this issue?

Thanks

Asalih3d

Without having too much knowledge about the topic, I suppose it could be a "Nested Loop" which is going wrong. Maybe some memory issue was experienced because of stability issues or overload on the GPU, if some other software was running too. If it is a single phase simulation you should be able to restart with help from the docs, else maybe try again with a rougher dp to see if the error occurs at the same time.

Just a few suggestions, I really do not know too much about that part.

Kind regards

Alex

Sometimes those errors appear when GPU features are not enough or if you are using a GPU with compute capability lower than the one for which the code was compiled... Please can you tell us what GPU card (and compute capability) are you using as execution device?

Regards

aoliva

Thank you Alex. Here below the information about my GPU card

Device 0: "GeForce GTX 960"
CUDA Capability Major....: 5.2
Global memory............: 4042 MBytes
CUDA Cores...............: 1024 ( 8 Multiprocessors, 128 CUDA Cores/MP)
GPU Max Clock rate.......: 1329 MHz (1.33 GHz)
Memory Clock rate........: 3505 Mhz
Memory Bus Width.........: 128-bit
L2 Cache Size............: 1024 KBytes
Constant memory..........: 64 KBytes
Shared memory per block..: 48 KBytes
Registers per block......: 65536
Maximum threads per MP...: 2048
Maximum threads per block: 1024
Concurrent copy and kernel execution....: Yes with 2 copy engine(s)
Run time limit on kernels...............: Yes
Integrated GPU sharing Host Memory......: No
Support host page-locked memory mapping.: Yes
Device has ECC support..................: Disabled
Device supports Unified Addressing (UVA): Yes
Device PCI (Domain / Bus / location)....: 0 / 15 / 0
Device supports P2P and RDMA............: Yes
Device supports P2P from/to GPUs........:

[GPU Hardware]
Device default: 0 "GeForce GTX 960"
Compute capability: 5.2
Memory global: 4042 MB
Memory shared: 49152 Bytes

Regards,