Table 2.
These tiling parameter sets were found to provide optimum performance under the conditions described in Tab. 1. Also presented is the amount of shared memory used per block (Mem./B.) for each set.
Set | Nblock | Nbin | Nconst | Ngrid | Mem./B. |
---|---|---|---|---|---|
cc1.0_8192 | 32 | 1024 | 5440 | 256 | 4.38 kB |
cc1.2_8192_a | 320 | 3072 | 5440 | 256 | 15.75 kB |
cc1.2_1024_a | 256 | 1024 | 5440 | 512 | 7.00 kB |
cc2.0_8192_a | 896 | 8192 | 5440 | 256 | 42.50 kB |