pub fn prewarm(stream: u64)Expand description
2026-09-25: Run one 64x64x64 BF16 GEMM on stream at model load
(serve_load.rs), so the first cuBLASLt call, including ctx’s
cublasLtCreate and workspace allocation, happens before the first
request.
Returns nothing: a failure is logged as a warning, and the first GEMM of a request then pays the same costs.