zgemvi_buffer_size# nvmath.bindings.cusparse.zgemvi_buffer_size( intptr_t handle, int trans_a, int m, int n, int nnz, ) → int[source]# See cusparseZgemvi_bufferSize.
