Hi,
We use eight Tesla P100 to run germline pipeline with NA12878 data, But getting the following error:
|| Parabricks accelerated Genomics Pipeline ||
|| Version v2.5.0 ||
|| GPU-BWA mem, Sorting, Marking Duplicates, BQSR ||
|| Contact: Parabricks-Support@nvidia.com ||
[M::bwa_idx_load_from_disk] read 0 ALT contigs
GPU-BWA mem
ProgressMeter Reads Base Pairs Aligned
[09:55:00] 5061412 750000000
[09:55:22] 10122400 1510000000
[09:55:44] 15182642 2280000000
[09:56:07] 20242532 3030000000
[09:56:30] 25301622 3810000000
[09:56:53] 30361222 4570000000
[09:57:15] 35421664 5320000000
[09:57:37] 40481834 6090000000
[09:58:00] 45541066 6830000000
[09:58:24] 50599860 7600000000
[09:58:46] 55658706 8350000000
[09:59:09] 60719092 9090000000
cudaSafeCall() failed at ParaBricks/src/mem_chain_kernel.cu:3920 : an illegal memory access was encountered
GPU: 3, cudaSafeCall() failed at ParaBricks/src/mem_chain_kernel.cu:3889 : an illegal memory access was encountered
GPU: 0, cudaSafeCall() failed at ParaBricks/src/mem_chain2aln_kernel.cu:5145 : an illegal memory access was encountered
GPU: 1, cudaSafeCall() failed at ParaBricks/src/mem_chain2aln_kernel.cu:5145 : an illegal memory access was encountered
Any suggestions or advice will be appreciated