aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2019-11-02Adapt to upstream changesAdrian Kummerlaender
2019-10-30Rename folder, add basic README.mdAdrian Kummerlaender
2019-10-30Extract settings into config file, add documentationAdrian Kummerlaender
2019-10-30Move C++ LDC template from upstream, improve buildAdrian Kummerlaender
2019-10-29Merge shell environmentsAdrian Kummerlaender
2019-10-29Add example for C++ targetAdrian Kummerlaender
2019-10-29Move example into subfolderAdrian Kummerlaender
2019-10-29Adapt to upstream changesAdrian Kummerlaender
2019-10-29Fix double precisionAdrian Kummerlaender
2019-10-28Basic 2D LDC using boltzgen for kernel generationAdrian Kummerlaender
Using cell lists as parameters for multiple non-branching kernels seems to reduce performance by ~50 MLUPS (for single precision D2Q9). This might be alleviated by padding the cell lists to enable thread layout control or by improved kernel dispatching. On the upside this OpenCL program runs not only on GPUs but is also vectorized on Intel CPUs yielding about 180 MLUPS (single precision) and - anticlimactically - 85 MLUPS for double precision on a i7-4790K. However both these values compare well to the performance of established CPU LBM codes.