Index

_

__init__() (triton.Config method) (triton.experimental.gluon.GluonJITFunction method) (triton.experimental.gluon.language.amd.AMDMFMALayout method) (triton.experimental.gluon.language.amd.AMDWMMALayout method) (triton.experimental.gluon.language.amd.warp_pipeline_stage method) (triton.experimental.gluon.language.AutoLayout method) (triton.experimental.gluon.language.BlockedLayout method) (triton.experimental.gluon.language.CoalescedLayout method) (triton.experimental.gluon.language.distributed_type method) (triton.experimental.gluon.language.DistributedLinearLayout method) (triton.experimental.gluon.language.DotOperandLayout method) (triton.experimental.gluon.language.nvidia.ampere.mbarrier.MBarrierLayout method) (triton.experimental.gluon.language.nvidia.blackwell.clc.clc_result method) (triton.experimental.gluon.language.nvidia.blackwell.tensor_memory_descriptor method) (triton.experimental.gluon.language.nvidia.blackwell.tensor_memory_descriptor_type method) (triton.experimental.gluon.language.nvidia.blackwell.TensorMemoryLayout method) (triton.experimental.gluon.language.nvidia.blackwell.TensorMemoryScalesLayout method) (triton.experimental.gluon.language.nvidia.blackwell.tma.tensor_descriptor method) (triton.experimental.gluon.language.nvidia.blackwell.tma.tensor_descriptor_type method) (triton.experimental.gluon.language.nvidia.hopper.mbarrier.MBarrierLayout method) (triton.experimental.gluon.language.nvidia.hopper.tma.tensor_descriptor method) (triton.experimental.gluon.language.nvidia.hopper.tma.tensor_descriptor_im2col method) (triton.experimental.gluon.language.nvidia.hopper.tma.tensor_descriptor_im2col_type method) (triton.experimental.gluon.language.nvidia.hopper.tma.tensor_descriptor_type method) (triton.experimental.gluon.language.NVMMADistributedLayout method) (triton.experimental.gluon.language.NVMMASharedLayout method) (triton.experimental.gluon.language.PaddedSharedLayout method) (triton.experimental.gluon.language.shared_memory_descriptor method) (triton.experimental.gluon.language.SharedLinearLayout method) (triton.experimental.gluon.language.SliceLayout method) (triton.experimental.gluon.language.static_range method) (triton.experimental.gluon.language.SwizzledSharedLayout method) (triton.experimental.gluon.language.tensor method) (triton.experimental.gluon.nvidia.blackwell.TensorDescriptor method) (triton.experimental.gluon.nvidia.hopper.TensorDescriptor method) (triton.experimental.gluon.nvidia.hopper.TensorDescriptorIm2Col method) (triton.language.range method) (triton.language.static_range method) (triton.language.tensor method) (triton.language.tensor_descriptor method) (triton.testing.Benchmark method)

A

abs() (in module triton.experimental.gluon.language) (in module triton.language) add() (in module triton.experimental.gluon.language) advance() (in module triton.language) allocate_mbarrier() (in module triton.experimental.gluon.language.nvidia.ampere.mbarrier) (in module triton.experimental.gluon.language.nvidia.hopper.mbarrier) allocate_shared_memory() (in module triton.experimental.gluon.language) allocate_tensor_memory() (in module triton.experimental.gluon.language.nvidia.blackwell) AMDMFMALayout (class in triton.experimental.gluon.language.amd) AMDWMMALayout (class in triton.experimental.gluon.language.amd) arange() (in module triton.experimental.gluon.language) (in module triton.language) argmax() (in module triton.language) argmin() (in module triton.language) arrive() (in module triton.experimental.gluon.language.nvidia.ampere.mbarrier) (in module triton.experimental.gluon.language.nvidia.hopper.cluster) (in module triton.experimental.gluon.language.nvidia.hopper.mbarrier) assert_close() (in module triton.testing) associative_scan() (in module triton.experimental.gluon.language) (in module triton.language) assume() (in module triton.experimental.gluon.language) (in module triton.language) async_atomic_add() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_atomic_and() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_atomic_max() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_atomic_min() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_atomic_or() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_atomic_xor() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma)

async_copy_global_to_shared() (in module triton.experimental.gluon.language.nvidia.ampere.async_copy) (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_copy_global_to_shared_im2col() (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_copy_shared_to_global() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_gather() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) async_load() (in module triton.experimental.gluon.language.nvidia.ampere.async_copy) (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_load_im2col() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) async_scatter() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) async_store() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) atomic_add() (in module triton.experimental.gluon.language) (in module triton.language) atomic_and() (in module triton.experimental.gluon.language) (in module triton.language) atomic_cas() (in module triton.experimental.gluon.language) (in module triton.language) atomic_max() (in module triton.experimental.gluon.language) (in module triton.language) atomic_min() (in module triton.experimental.gluon.language) (in module triton.language) atomic_or() (in module triton.experimental.gluon.language) (in module triton.language) atomic_xchg() (in module triton.experimental.gluon.language) (in module triton.language) atomic_xor() (in module triton.experimental.gluon.language) (in module triton.language) AutoLayout (class in triton.experimental.gluon.language) autotune() (in module triton)

B

bank_conflicts() (in module triton.experimental.gluon.language) barrier() (in module triton.experimental.gluon.language) (in module triton.experimental.gluon.language.nvidia.hopper.cluster) Benchmark (class in triton.testing) BlockedLayout (class in triton.experimental.gluon.language) broadcast() (in module triton.experimental.gluon.language) (in module triton.language) broadcast_to() (in module triton.language) buffer_atomic_add() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) buffer_atomic_and() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) buffer_atomic_max() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4)

buffer_atomic_min() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) buffer_atomic_or() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) buffer_atomic_xchg() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) buffer_atomic_xor() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) buffer_load() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) buffer_store() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4)

C

cast() (in module triton.experimental.gluon.language) (in module triton.language) cat() (in module triton.language) cdiv() (in module triton.experimental.gluon.language) (in module triton.language) ceil() (in module triton.experimental.gluon.language) (in module triton.language) clamp() (in module triton.experimental.gluon.language) (in module triton.language)

clc_result (class in triton.experimental.gluon.language.nvidia.blackwell.clc) CoalescedLayout (class in triton.experimental.gluon.language) commit_group() (in module triton.experimental.gluon.language.nvidia.ampere.async_copy) Config (class in triton) constexpr_function() (in module triton.experimental.gluon) convert_layout() (in module triton.experimental.gluon.language) cos() (in module triton.experimental.gluon.language) (in module triton.language) cumprod() (in module triton.language) cumsum() (in module triton.language)

D

debug_barrier() (in module triton.language) device_assert() (in module triton.experimental.gluon.language) (in module triton.language) device_print() (in module triton.experimental.gluon.language) (in module triton.language) distributed_type (class in triton.experimental.gluon.language) DistributedLinearLayout (class in triton.experimental.gluon.language) div_rn() (in module triton.experimental.gluon.language) (in module triton.language)

do_bench() (in module triton.testing) do_bench_cudagraph() (in module triton.testing) do_bench_cudagraph_proton() (in module triton.testing) do_bench_proton() (in module triton.testing) dot() (in module triton.language) dot_fma() (in module triton.experimental.gluon.language) dot_scaled() (in module triton.language) DotOperandLayout (class in triton.experimental.gluon.language)

E

erf() (in module triton.experimental.gluon.language) (in module triton.language) exp() (in module triton.experimental.gluon.language) (in module triton.language)

exp2() (in module triton.experimental.gluon.language) (in module triton.language) expand_dims() (in module triton.experimental.gluon.language) (in module triton.language) expect() (in module triton.experimental.gluon.language.nvidia.hopper.mbarrier)

F

fdiv() (in module triton.experimental.gluon.language) (in module triton.language) fence_async_shared() (in module triton.experimental.gluon.language.nvidia.blackwell) (in module triton.experimental.gluon.language.nvidia.hopper) fence_init_release_cluster() (in module triton.experimental.gluon.language.nvidia.hopper.mbarrier) flip() (in module triton.language) floor() (in module triton.experimental.gluon.language) (in module triton.language)

fma() (in module triton.experimental.gluon.language) (in module triton.language) fp4_to_fp() (in module triton.experimental.gluon.language) full() (in module triton.experimental.gluon.language) (in module triton.language) full_like() (in module triton.experimental.gluon.language)

G

gather() (in module triton.experimental.gluon.language) (in module triton.language) gdc_launch_dependents() (in module triton.language.extra.cuda)

gdc_wait() (in module triton.language.extra.cuda) get_mfma_scale_layout() (in module triton.experimental.gluon.language.amd.cdna4) GluonJITFunction (class in triton.experimental.gluon)

H

heuristics() (in module triton)

histogram() (in module triton.experimental.gluon.language) (in module triton.language)

I

init() (in module triton.experimental.gluon.language.nvidia.ampere.mbarrier) (in module triton.experimental.gluon.language.nvidia.hopper.mbarrier) inline_asm_elementwise() (in module triton.experimental.gluon.language) (in module triton.language)

interleave() (in module triton.language) invalidate() (in module triton.experimental.gluon.language.nvidia.ampere.mbarrier) (in module triton.experimental.gluon.language.nvidia.hopper.mbarrier)

J

jit() (in module triton) (in module triton.experimental.gluon)

join() (in module triton.experimental.gluon.language) (in module triton.language)

L

load() (in module triton.experimental.gluon.language) (in module triton.language) load_result() (in module triton.experimental.gluon.language.nvidia.blackwell.clc) load_tensor_descriptor() (in module triton.language)

log() (in module triton.experimental.gluon.language) (in module triton.language) log2() (in module triton.experimental.gluon.language) (in module triton.language)

M

make_block_ptr() (in module triton.language) make_tensor_descriptor() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) (in module triton.language) map_elementwise() (in module triton.experimental.gluon.language) max() (in module triton.experimental.gluon.language) (in module triton.language) max_constancy() (in module triton.experimental.gluon.language) (in module triton.language) max_contiguous() (in module triton.experimental.gluon.language) (in module triton.language) maximum() (in module triton.experimental.gluon.language) (in module triton.language) mbarrier_arrive() (in module triton.experimental.gluon.language.nvidia.ampere.async_copy) MBarrierLayout (class in triton.experimental.gluon.language.nvidia.ampere.mbarrier) (class in triton.experimental.gluon.language.nvidia.hopper.mbarrier) mfma() (in module triton.experimental.gluon.language.amd.cdna3) (in module triton.experimental.gluon.language.amd.cdna4) mfma_scaled() (in module triton.experimental.gluon.language.amd.cdna4) min() (in module triton.experimental.gluon.language) (in module triton.language)

minimum() (in module triton.experimental.gluon.language) (in module triton.language) mma_v2() (in module triton.experimental.gluon.language.nvidia.ampere) (in module triton.experimental.gluon.language.nvidia.blackwell) (in module triton.experimental.gluon.language.nvidia.hopper) module triton.experimental.gluon.language.amd.cdna4.async_copy triton.experimental.gluon.language.nvidia.ampere.async_copy triton.experimental.gluon.language.nvidia.ampere.mbarrier triton.experimental.gluon.language.nvidia.blackwell.async_copy triton.experimental.gluon.language.nvidia.blackwell.clc triton.experimental.gluon.language.nvidia.blackwell.mbarrier triton.experimental.gluon.language.nvidia.blackwell.tma triton.experimental.gluon.language.nvidia.hopper.async_copy triton.experimental.gluon.language.nvidia.hopper.cluster triton.experimental.gluon.language.nvidia.hopper.mbarrier triton.experimental.gluon.language.nvidia.hopper.tma mul() (in module triton.experimental.gluon.language) multiple_of() (in module triton.experimental.gluon.language) (in module triton.language) must_use_result() (in module triton.experimental.gluon)

N

num_ctas() (in module triton.experimental.gluon.language) num_programs() (in module triton.experimental.gluon.language) (in module triton.language)

num_warps() (in module triton.experimental.gluon.language) NVMMADistributedLayout (class in triton.experimental.gluon.language) NVMMASharedLayout (class in triton.experimental.gluon.language)

P

PaddedSharedLayout (class in triton.experimental.gluon.language) perf_report() (in module triton.testing) permute() (in module triton.experimental.gluon.language) (in module triton.language)

program_id() (in module triton.experimental.gluon.language) (in module triton.language)

R

rand() (in module triton.language) randint() (in module triton.language) randint4x() (in module triton.language) randn() (in module triton.language) range (class in triton.language) ravel() (in module triton.experimental.gluon.language) (in module triton.language)

reduce() (in module triton.experimental.gluon.language) (in module triton.language) reduce_or() (in module triton.experimental.gluon.language) reshape() (in module triton.experimental.gluon.language) (in module triton.language) rsqrt() (in module triton.experimental.gluon.language) (in module triton.language)

S

set_auto_layout() (in module triton.experimental.gluon.language) shared_memory_descriptor (class in triton.experimental.gluon.language) SharedLinearLayout (class in triton.experimental.gluon.language) sigmoid() (in module triton.language) sin() (in module triton.experimental.gluon.language) (in module triton.language) SliceLayout (class in triton.experimental.gluon.language) softmax() (in module triton.language) sort() (in module triton.language) split() (in module triton.experimental.gluon.language) (in module triton.language) sqrt() (in module triton.experimental.gluon.language) (in module triton.language) sqrt_rn() (in module triton.experimental.gluon.language) (in module triton.language)

static_assert() (in module triton.experimental.gluon.language) (in module triton.language) static_print() (in module triton.experimental.gluon.language) (in module triton.language) static_range (class in triton.experimental.gluon.language) (class in triton.language) store() (in module triton.experimental.gluon.language) (in module triton.language) store_tensor_descriptor() (in module triton.language) store_wait() (in module triton.experimental.gluon.language.nvidia.blackwell.tma) (in module triton.experimental.gluon.language.nvidia.hopper.tma) sub() (in module triton.experimental.gluon.language) sum() (in module triton.experimental.gluon.language) (in module triton.language) swizzle2d() (in module triton.language) SwizzledSharedLayout (class in triton.experimental.gluon.language)

T

tensor (class in triton.experimental.gluon.language) (class in triton.language) tensor_descriptor (class in triton.experimental.gluon.language.nvidia.blackwell.tma) (class in triton.experimental.gluon.language.nvidia.hopper.tma) (class in triton.language) tensor_descriptor_im2col (class in triton.experimental.gluon.language.nvidia.hopper.tma) tensor_descriptor_im2col_type (class in triton.experimental.gluon.language.nvidia.hopper.tma) tensor_descriptor_type (class in triton.experimental.gluon.language.nvidia.blackwell.tma) (class in triton.experimental.gluon.language.nvidia.hopper.tma) tensor_memory_descriptor (class in triton.experimental.gluon.language.nvidia.blackwell) tensor_memory_descriptor_type (class in triton.experimental.gluon.language.nvidia.blackwell) TensorDescriptor (class in triton.experimental.gluon.nvidia.blackwell) (class in triton.experimental.gluon.nvidia.hopper) TensorDescriptorIm2Col (class in triton.experimental.gluon.nvidia.hopper) TensorMemoryLayout (class in triton.experimental.gluon.language.nvidia.blackwell) TensorMemoryScalesLayout (class in triton.experimental.gluon.language.nvidia.blackwell) to_linear_layout() (in module triton.experimental.gluon.language) to_tensor() (in module triton.experimental.gluon.language) topk() (in module triton.language) trans() (in module triton.language) triton.experimental.gluon.language.amd.cdna4.async_copy module

triton.experimental.gluon.language.nvidia.ampere.async_copy module triton.experimental.gluon.language.nvidia.ampere.mbarrier module triton.experimental.gluon.language.nvidia.blackwell.async_copy module triton.experimental.gluon.language.nvidia.blackwell.clc module triton.experimental.gluon.language.nvidia.blackwell.mbarrier module triton.experimental.gluon.language.nvidia.blackwell.tma module triton.experimental.gluon.language.nvidia.hopper.async_copy module triton.experimental.gluon.language.nvidia.hopper.cluster module triton.experimental.gluon.language.nvidia.hopper.mbarrier module triton.experimental.gluon.language.nvidia.hopper.tma module try_cancel() (in module triton.experimental.gluon.language.nvidia.blackwell.clc)

U

umulhi() (in module triton.experimental.gluon.language) (in module triton.language)

V

view() (in module triton.language)

W

wait() (in module triton.experimental.gluon.language.nvidia.ampere.mbarrier) (in module triton.experimental.gluon.language.nvidia.hopper.cluster) (in module triton.experimental.gluon.language.nvidia.hopper.mbarrier) wait_group() (in module triton.experimental.gluon.language.nvidia.ampere.async_copy) warp_pipeline_stage (class in triton.experimental.gluon.language.amd) warp_specialize() (in module triton.experimental.gluon.language)

warpgroup_mma() (in module triton.experimental.gluon.language.nvidia.hopper) warpgroup_mma_wait() (in module triton.experimental.gluon.language.nvidia.hopper) where() (in module triton.experimental.gluon.language) (in module triton.language) wmma() (in module triton.experimental.gluon.language.amd.rdna3) (in module triton.experimental.gluon.language.amd.rdna4)

X

xor_sum() (in module triton.experimental.gluon.language) (in module triton.language)

Z

zeros() (in module triton.experimental.gluon.language) (in module triton.language)

zeros_like() (in module triton.experimental.gluon.language) (in module triton.language)