Home
last modified time | relevance | path

Searched refs:warp_size (Results 1 – 1 of 1) sorted by relevance

/libCEED/backends/cuda-gen/
H A Dceed-cuda-gen-operator.c40 static int Waste(int threads_per_sm, int warp_size, int threads_per_elem, int elems_per_block) { in Waste() argument
43 int block_size = CeedDivUpInt(useful_threads_per_block, warp_size) * warp_size; in Waste()
75 …elem, int blocks_per_sm, int max_threads_per_block, int max_threads_z, int warp_size, int block[3], in BlockGridCalculate() argument
80 int waste = Waste(threads_per_sm, warp_size, threads_per_elem, 1); in BlockGridCalculate()
83 int i_waste = Waste(threads_per_sm, warp_size, threads_per_elem, i); in BlockGridCalculate()