pub const W4A16_GEMV_OUTS_PER_BLOCK: u32 = 4;
2026-09-25: Outputs per 256-thread block of w4a16_gemv; equals #define N_PER_BLOCK in kernels/gb10/common/w4a16_gemv.cu.
w4a16_gemv
#define N_PER_BLOCK