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 struct
DecodeOutProducer
struct DecodeOutProducer[dtype: DType, config: MLA_SM100_Decode_Config]
Fields
- pipe (
DecodeOutProducer[dtype, config].OutPipeType): - smem (
UnsafePointer[Scalar[dtype], MutAnyOrigin, address_space=AddressSpace.SHARED]):
Implemented traits
AnyType,
Copyable,
ImplicitlyCopyable,
ImplicitlyDeletable,
Movable,
RegisterPassable,
TrivialRegisterPassable
comptime members
block_per_warp
comptime block_per_warp = (DecodeOutProducer[dtype, config].col_per_warp // (config // 4))
blocks_per_stage
comptime blocks_per_stage = 2 if (DecodeOutProducer[dtype, config].block_per_warp != 0) else 1
col_per_warp
comptime col_per_warp = (config // 2)
num_out_blocks
comptime num_out_blocks = (config // (config // 4))
num_out_stages
comptime num_out_stages = (DecodeOutProducer[dtype, config].num_out_blocks // DecodeOutProducer[dtype, config].blocks_per_stage)
out_stage_bytes
comptime out_stage_bytes = (DecodeOutProducer[dtype, config].out_stage_elems * size_of[dtype]())
out_stage_elems
comptime out_stage_elems = (config * (config // 4))
OutPipeType
comptime OutPipeType = OutPipeline[DecodeOutProducer[dtype, config].num_out_stages, WARPGROUP_SIZE, 1]
Methods
__init__
def __init__(pipe: OutPipeline[Self.num_out_stages, WARPGROUP_SIZE, 1], smem: UnsafePointer[Scalar[dtype], MutAnyOrigin, address_space=AddressSpace.SHARED]) -> Self
init
def init(self)
stage_base_ptr
def stage_base_ptr(self, half_idx: Int) -> UnsafePointer[Scalar[dtype], MutAnyOrigin, address_space=AddressSpace.SHARED]
Returns:
UnsafePointer[Scalar[dtype], MutAnyOrigin, address_space=AddressSpace.SHARED]
producer_mbar
def producer_mbar(self) -> MBarType
Returns:
MBarType
acquire
def acquire(self)
commit_step
def commit_step(mut self)
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!