NVFP4_GROUPED_TC_ROWS_PER_PASS

Constant NVFP4_GROUPED_TC_ROWS_PER_PASS 

Source
pub const NVFP4_GROUPED_TC_ROWS_PER_PASS: u32 = 8;
Expand description

2026-10-02: Rows per pass of the tensor-core kernels (moe_nvfp4_grouped_tc.cu, TC_ROWS), gate+up and down alike.