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

ep

Expert Parallelism (EP) Communication Kernel.

comptime values

RT_LAYOUT_2D

comptime RT_LAYOUT_2D = Layout[*?, *?]

Structs

• ​DistributedEPCombine: Registers the mo.distributed.ep.combine graph op with the graph compiler.
• ​DistributedEPDispatch: Registers the mo.distributed.ep.dispatch graph op with the graph compiler.
• ​DistributedEPDispatchBlockScaledNV: Registers the mo.distributed.ep.dispatch.block.scaled.nv graph op with the graph compiler.
• ​DistributedEPDispatchFP8: Registers the mo.distributed.ep.dispatch.fp8 graph op with the graph compiler.
• ​DistributedEPDispatchMXFP4: Registers the mo.distributed.ep.dispatch.mxfp4 graph op with the graph compiler.
• ​Struct_ep_combine: Registers the ep.combine graph op with the graph compiler.
• ​Struct_ep_combine_async: Registers the ep.combine_async graph op with the graph compiler.
• ​Struct_ep_combine_skip_a2a: Registers the ep.combine.skip_a2a graph op with the graph compiler.
• ​Struct_ep_combine_wait: Registers the ep.combine_wait graph op with the graph compiler.
• ​Struct_ep_dispatch: Registers the ep.dispatch graph op with the graph compiler.
• ​Struct_ep_dispatch_async: Registers the ep.dispatch_async graph op with the graph compiler.
• ​Struct_ep_dispatch_async_block_scaled_nv: Registers the ep.dispatch_async.block.scaled.nv graph op with the graph compiler.
• ​Struct_ep_dispatch_async_mxfp4: Registers the ep.dispatch_async.mxfp4 graph op with the graph compiler.
• ​Struct_ep_dispatch_block_scaled_nv: Registers the ep.dispatch.block.scaled.nv graph op with the graph compiler.
• ​Struct_ep_dispatch_fp8: Registers the ep.dispatch.fp8 graph op with the graph compiler.
• ​Struct_ep_dispatch_mxfp4: Registers the ep.dispatch.mxfp4 graph op with the graph compiler.
• ​Struct_ep_dispatch_wait: Registers the ep.dispatch_wait graph op with the graph compiler.
• ​Struct_ep_dispatch_wait_block_scaled_nv: Registers the ep.dispatch_wait.block.scaled.nv graph op with the graph compiler.
• ​Struct_ep_dispatch_wait_fp8: Registers the ep.dispatch_wait.fp8 graph op with the graph compiler.
• ​Struct_ep_dispatch_wait_mxfp4: Registers the ep.dispatch_wait.mxfp4 graph op with the graph compiler.
• ​Struct_ep_fused_silu: Registers the ep.fused_silu graph op with the graph compiler.
• ​Struct_ep_fused_silu_fp8: Registers the ep.fused_silu.fp8 graph op with the graph compiler.
• ​Struct_ep_fused_silu_mxfp4: Registers the ep.fused_silu.mxfp4 graph op with the graph compiler.
• ​Struct_ep_fused_silu_nvfp4: Registers the ep.fused_silu.nvfp4 graph op with the graph compiler.
• ​Struct_ep_init: Registers the ep.init graph op with the graph compiler.

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