IMPORTANT: To view this page as Markdown, append `.md` to the URL (e.g. /max/get-started.md). For the complete documentation index, see llms.txt.
Skip to main content
For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo module

mla_decode_kv_bf16

Implements the SM100 (Blackwell) MLA decode kernel operating on a BF16 key-value cache.

Specializes the Multi-Latent Attention decode path for Blackwell (SM100) hardware, dispatching three warpgroups (softmax, correction, and load/MMA/store) that cooperate over shared-memory barriers and the TCGEN05 tensor-memory pipeline. The MLA_SM100_Decode_KV_BF16 struct carries the comptime configuration derived from MLA_SM100_Decode_Config and exposes the entry-point kernel along with the load, mmaQK, and mmaPV helper methods invoked by the individual warps.

Structs

  • MLA_SM100_Decode_KV_BF16: Implements the SM100 MLA decode kernel over a BF16 paged KV cache, packaging the kernel entry point with its load and MMA helpers.

Was this page helpful?