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_prefill_blockscale

Implements the SM100 (Blackwell) MLA prefill attention kernel with blockwise FP8 scaling, converting block-scaled FP8 K_rope tiles to BF16 in shared memory before the MMA stage.

Structs

• ​WarpRole: Tags each warp with its specialized role in the warp-specialized MLA prefill kernel.

Functions

• ​mla_sm100_prefill_blockscale: Launches the SM100 MLA prefill kernel with blockwise FP8 scaling.
• ​warp_idx_to_role: Maps a global warp index to its WarpRole based on warpgroup assignment.

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit