Function bf16_gemm_act_weight_t_f32_out 

pub fn bf16_gemm_act_weight_t_f32_out(
    act: u64,
    weight: u64,
    out: u64,
    m: u32,
    n: u32,
    k: u32,
    stream: u64,
) -> Result<()>

bf16_gemm_act_weight_t with BF16 inputs and an FP32 output buffer.

Exists for consumers whose downstream kernel reads FP32 and which therefore could not use the batched BF16-out path at all. GLM’s DSA indexer is the motivating case: its wq_b projection writes FP32, so select_rows_batched fell back to one M=1 GEMV PER ROW — 256 full sweeps of the same weight per DSA layer per prefill chunk — under a comment reading “no FP32-out batchm twin exists”. This is that twin.

Identical to the BF16 form except the D layout: cuBLASLt accumulates in FP32 either way (CUBLAS_COMPUTE_32F), so this actually stores MORE of the accumulator than the BF16 output does, rather than less.