the chunked openCL execution waits for 16 chunks to finish until it queues the next ones, so that the system gets some time to check whether the GPU still responds
Signed-off-by: Sattler <sattler@ad.dkfz-heidelberg.de>
Description
Description
Details
Details
- Provenance
• sattler Authored on Dec 8 2017, 1:08 PM • sattler Pushed on Dec 8 2017, 1:08 PM - Parents
- rMITKb343b1cb4e5b: introduced chunked calculation for the UsedLinesCalculation
- Branches
- Unknown
- Tags