pystencils.backend.platforms.CudaPlatform#
- class pystencils.backend.platforms.CudaPlatform(ctx, indexing_rank, default_block_size, *, assume_warp_aligned_block_size=False, warp_size=None, use_cub_reductions=False)#
Platform for the CUDA GPU target.
- Parameters:
ctx (KernelCreationContext)
indexing_rank (int)
default_block_size (dim3 | None)
assume_warp_aligned_block_size (bool)
warp_size (int | None)
use_cub_reductions (bool)