gemm_buffer_size

nvmath.bindings.cublasMp.gemm_buffer_size(

intptr_t handle,

int trans_a,

int trans_b,

int64_t m,

int64_t n,

int64_t k,

intptr_t alpha,

intptr_t a,

int64_t ia,

int64_t ja,

intptr_t desc_a,

intptr_t b,

int64_t ib,

int64_t jb,

intptr_t desc_b,

intptr_t beta,

intptr_t c,

int64_t ic,

int64_t jc,

intptr_t desc_c,

int compute_type,

) → tuple

See cublasMpGemm_bufferSize.