Commit Graph

Select branches

Hide Pull Requests

abstractions_4

add_local_buffer

after_qol

algebraic_upat

all_bv_no_idx

all_sink

amd_isel

amd_refac

amd_smu_reset_patch

amd_sqtt

amd_uop

amd_uop_cleanups

assign_copy

assign_try_3

before2

better_spec

better_syn

bitcast_spec

broadcast_shape_2

bt_ei

buf_heur

bump_nv_2404

cache_reshape

call_inline

call_is_kernels

callify

cast_vec_as_cast

cat_mop

changes_from_rangify

check_clsetarg

check_dims

cifar_30

cifar_30_clean

clean_load

clean_spec

clone_tg

codegen_try_2

contig_on_after

continue_mat

deepsek

default_late_allreduce

delete_kernel_py

delete_ones

delete_slow_rangeify

delta_attention

delta_attention_scan

demote_op

disable_rules

dsp_search

dsp_search_merged

dtype_shape

earlier_gater

emu_cdna

emulate_ctx

end_all_ranges

endrange

er_prioity

execution_unit

failing_assign

fancy_usb

fast_llm

fix_fa_fwd_clang_22

fix_fuse

fix_llvm_cast

fix_range_merging_sd

fix_rangeify_tests

fix_rc

fix_sched_copy

fix_slice_store

fix_src_spec

fixup_fix_kernel

full_spec

fun_w_egraphs

gate-on-load-store

gate_rewrite

gemma4_gpt

gen_pdf_fast

gh-pages

gl_dims

gldims

glm_flash

grad_outerworld

grok_x86_novec

group_div_rules

hcq2_vram_copies

heur_4096

heur_group

hk_gemm

http_threading

hwtest_fixes

id_store

if_range_cleanups

ignore_oob

image_idx_xy

image_in_mc

image_no_vec

image_refactors_stack

index_to_slice

invalid_try_3

jinja_template

kernel4

kernel_is_call

kernelless

kimi_slop

kitten_matmul

late_Locals

late_add_load

late_allreduce

late_codegen

late_decomp_is_late

late_lower_idx

late_view

late_weakint

lds_double_buffering

less_wmma_arg

lil_dg_change

lil_rangeify_td

llama_trainer

llm_262k_minimal

llm_flash_attn

llm_qwen

llm_serve_fixes

load_view_late

local_cache

locals_have_device

loop_end_op

loop_is_rng

lowerer_hacks

mac_pytest

manual_kernel_3

master

mega_lowerer

min_new_lin

min_outer_range

minor_views

mixins7

mlperf_training_v4.1

mlperf_training_v5.0

mlperf_training_v6.0

modern_scan

more_apps

more_early_comps

more_folding_rules

more_kernelcount

more_pcode

more_rangeify_pm_tests

more_ren_clean

more_uop_gc_test

more_uop_mm

more_vmap

move_devec

move_dtype_decomps

move_expander

move_gates_to_load_store

move_mop_cleanup

move_schedule

move_simp_views

moveleftright

mt_typing

multi_range

multioutput

new_ck

new_devec

new_invalid_and_image

new_linearizer

new_shape

new_test

no_assign_2

no_buf_0

no_decimals

no_late_simplify_marg

no_merge_ends

no_merge_views

no_mstack_in_allreduce

no_one_ranges

no_sip_usbgpu

no_tsink_base

no_tuplize

no_user_ck_contig

no_vec_x86_4

no_weakint_variable

noop_index

novecimg

num_batches_tracked_shape

olmo3_support

one_compiler

one_hcopt

only_assign_buffers

only_loop_backends

only_reg_emu

only_reg_emu2

outer_range

outerworld_work

pad_invalid_work

pad_mixin

param_call_is_function

parse_sqtt_more

pci_iommu

port_nir

postrange_hcopts

precompile_backward

prefill_qwen_fast

print_special_ops

ps_clean

pyremu_try

python_prg

python_speed

q4k_support

qcom-openpilot-13ms

qcom_830

qcom_mmapeak

qcom_speed_codex

qol_string

qwen36_27b_amd_900

qwen36_27b_amd_minimal

qwen36_amd_bugfixes

qwen_mergable

r_cleanups

r_new_opt

ranged_call

rangeify_2

rangeify_const_folding

rdna3_vibes

rdna4_asm

rdna4_gemm

real_pyrender

recurse_sub

reg_coalese

relu_pattern

remove_assign

remove_cast_hack

remove_ck_contig

remove_define_var

remove_forced_re

remove_on_stack

remove_programspec

remove_pylint

remove_slow_tests

remove_vec_2

remove_vec_3

renderers_use_shapes_2

replace_if_with_range

reprocess_node_experiments

reshape_trait

resnet-ci

revert-11547-fix_mismatch_red

revert-11715-revert-11711-amd_llvm_tc_cast

revert-11809-no_ki_in_gpudims

revert-11811-better_div_nesting

revert-11876-memplan_vbig

revert-12104-simplify_valid_in_symbolic

revert-12308-bufferize_early

revert-12731-jitlink_init_check

revert-12813-after_cleanups

revert-13046-better_cleanup_arange_buffers

revert-13311-prof_link_sqtt

revert-13583-rocr_desc_2

revert-14734-tload

revert-17141-full_rewrite_to_sink

revert-17408-_broadcasted-weak

revert-9438-simple_matmul_np_rand

revert_wrong_ck_fix

rewrite_rangeify2

rtoposort

scan_assign

scan_op_8

sched_cache

sched_lin

sd_load_simple

self_type

set_testpath

shrink_in_render

simp_spec

simple_gidxs

simple_pool

simple_priority

simple_rangeify

simpler-sd-compile

simpler_end

simpler_fusion

simpler_image

simpler_postrange

simpler_reduce

slice_like_shrink

spec_3

spec_cleanup

spec_ren

speed_one_linear

split_buf_idx

split_decomp_pass

split_sub

stack_0

store_is_void

string_ranges

support_opts_in_contig

swizzle_in_kernel

sym_llm

sym_work

symbolic-chunk-split

szdiff_bot_test

tc_is_heur

tensor_graph_single_sink

test_amd_crash

test_fa

test_from_postopt

test_hgemm

test_unit

tiny_scan

tmr

tool_call_support

topoprop

ttomsa-new_linearizer

tuplegtuple

two_stage_remove

typed_checks

typing_work

unused_permute_arg

uop_prg

upd_test4

update_benchmark

update_benchmark_search

update_mlperf

user_failure

vecless_x86_glm

verify_sink_dims_clean

vf_ish

view_after_const

view_in_codegen

viz_cbroadcast

warp_fun

wgpu-f16-packed

working_postopt

x86_moves

x86_numel

v0.10.0

v0.10.1

v0.10.2

v0.10.3

v0.11.0

v0.12.0

v0.13.0

v0.4.0

v0.5.0

v0.6.0

v0.7.0

v0.8.0

v0.9.0

v0.9.1

v0.9.2

Mono Color

• 90a5a312eb simplify ShapeTracker in UOp.const [pr] (#11424) chenyuandGitHub 2025-07-29 15:04:06 -04:00
• 398594029b spec checks arg of VIEW are ShapeTracker (#11422) chenyuandGitHub 2025-07-29 14:05:12 -04:00
• 1f1f99c287 hotfix: add DEBUG=3 to driver CI geohot 2025-07-29 11:03:47 -07:00
• 50fae54175 global local dims in gpudims [pr] (#11420) George HotzandGitHub 2025-07-29 10:39:03 -07:00
• 30b9c6ed26 global local dims in gpudims [pr] gldims geohot 2025-07-29 10:31:39 -07:00
• 9bc413f104 remove ShapeTracker.to_uop [pr] (#11418) chenyuandGitHub 2025-07-29 13:29:37 -04:00
• ba2c4df125 dont render cast ptrs standalone (#11417) George HotzandGitHub 2025-07-29 09:24:26 -07:00
• d38d285489 ci: add h machines (#11416) nimlgenandGitHub 2025-07-29 19:21:51 +03:00
• 2568bc0d99 ci: add caching for apt packages (#11162) 3 people 2025-07-29 18:04:56 +02:00
• 03909f2772 permute locals for HL uop matmul (#11412) George HotzandGitHub 2025-07-29 08:19:59 -07:00
• e0c9747684 amd: fix typo in has_scratch_base_registers for mi350 (#11413) nimlgenandGitHub 2025-07-29 10:30:06 +03:00
• 735ad5f10d kernel4 and 5 in uops (#11411) George HotzandGitHub 2025-07-28 19:35:48 -07:00
• c89fb841f8 cleanups kernel4 geohot 2025-07-28 19:22:35 -07:00
• 29f7d03d43 k5 support geohot 2025-07-28 19:11:27 -07:00
• b41e49472c kernel4 written in uops geohot 2025-07-28 17:02:24 -07:00
• 165d8e1263 k4 in python geohot 2025-07-28 16:56:14 -07:00
• 4b57aa2655 Revert "move simplify views to merge views" geohot 2025-07-28 16:19:51 -07:00
• ad1a2a68d5 add amd kernel 4 geohot 2025-07-28 16:14:11 -07:00
• 1e07dff384 move simplify views to merge views move_simp_views geohot 2025-07-28 13:55:23 -07:00
• fddc645668 HL=2 top matmul (#11406) George HotzandGitHub 2025-07-28 12:32:38 -07:00
• c7b4ab86e4 fix llvm tc on mi350 (#11404) nimlgenandGitHub 2025-07-28 21:37:43 +03:00
• 9f7c72ff8f remove UOp.valid method [pr] (#11402) chenyuandGitHub 2025-07-28 11:29:08 -04:00
• b22a34331b remove const valid in fixup_ast [pr] (#11401) chenyuandGitHub 2025-07-28 11:07:59 -04:00
• 7737cbb2a0 viz: tabulate runtime stats (#11400) qazalandGitHub 2025-07-28 20:56:39 +08:00
• 5e433685b1 reorder lowerer_hacks geohot 2025-07-27 19:21:08 -07:00
• 96c019a59c full gen geohot 2025-07-27 18:46:24 -07:00
• 938fe98daf more reduce into geohot 2025-07-27 18:34:52 -07:00
• ab6a27f627 remove a branch in UOp.r [pr] (#11398) chenyuandGitHub 2025-07-27 18:00:01 -04:00
• 052191eae4 Remote multihost (p2p with infiniband verbs) (#9746) 3 people 2025-07-28 02:44:32 +05:00
• 3ec3c27b83 hack on lowerer to make loops geohot 2025-07-27 11:55:26 -07:00
• a22417cc75 viz: fix bug with wrong program links (#11396) qazalandGitHub 2025-07-28 02:52:06 +08:00
• a5371f514b cpu: copies in profile (#11392) nimlgenandGitHub 2025-07-27 20:56:27 +03:00
• 8c10085459 assert shape on lowerer store [pr] (#11395) George HotzandGitHub 2025-07-27 10:41:57 -07:00
• 6174cfa828 viz: only show match counts greater than 0 (#11394) qazalandGitHub 2025-07-28 00:25:00 +08:00
• 3466a220de viz: disassembly viewer (#11393) qazalandGitHub 2025-07-27 23:44:28 +08:00
• 3bb232eb29 viz: query path in rewrite steps (#11391) qazalandGitHub 2025-07-27 19:51:47 +08:00
• b7ef73babd fix wmma ptx (#11389) 3 people 2025-07-27 14:28:35 +08:00
• 8dfcdb123d less wmma args (#11385) 4 people 2025-07-27 12:24:05 +08:00
• dfeee63d30 uop matmul work (#11388) George HotzandGitHub 2025-07-26 21:23:55 -07:00
• 3923e78061 no_vectorized_acc keeps single DEFINE_REG (#11387) George HotzandGitHub 2025-07-26 11:44:09 -07:00
• 4866ad57da viz: add runtime stats (#11383) qazalandGitHub 2025-07-27 01:40:46 +08:00
• 2c70eaf18c fix load / barrier (#11386) George HotzandGitHub 2025-07-26 10:27:37 -07:00
• 65673e68ca hcq: do not import during __del__ (#11384) nimlgenandGitHub 2025-07-26 13:58:55 +03:00
• 466ab5a3f2 store/load not pass through index (#11381) George HotzandGitHub 2025-07-25 21:01:47 -07:00
• 0a5f37946b unused permute arg on r (#11379) George HotzandGitHub 2025-07-25 19:52:37 -07:00
• 50e5c77814 sink PTRCAT store_is_void geohot 2025-07-25 19:52:25 -07:00
• be0c608b0a store is dtypes.void, no ptr pass through geohot 2025-07-25 19:43:14 -07:00
• d39b854a2d unused permute arg on r unused_permute_arg geohot 2025-07-25 19:31:47 -07:00
• 872ca998db simpler reduce that just removes ones simpler_reduce geohot 2025-07-25 19:17:38 -07:00
• 9a4f54a1f3 don't use the wmma args as much less_wmma_arg geohot 2025-07-25 18:46:46 -07:00
• e738b2d4a5 Merge branch 'master' into delete_ones delete_ones George HotzandGitHub 2025-07-25 18:28:23 -07:00
• 48562cb2db full shape simpler (#11376) George HotzandGitHub 2025-07-25 18:27:48 -07:00
• 3d68feb67d minor onnx Gather cleanup (#11375) chenyuandGitHub 2025-07-25 21:08:08 -04:00
• dfb3e99b09 late remove ones geohot 2025-07-25 15:51:56 -07:00
• d2473586d1 no keepdims in reduce geohot 2025-07-25 15:44:31 -07:00
• 88c338bfcc add kernelize to keccak for each data block (#11370) chenyuandGitHub 2025-07-25 16:07:20 -04:00
• dab07bcad9 use next instead of full list in UOp._device [pr] (#11369) chenyuandGitHub 2025-07-25 10:04:29 -04:00
• 1bb1f1aee8 hcq: fix race in _at_profile_finalize (#11368) nimlgenandGitHub 2025-07-25 14:14:02 +03:00
• 490a93902c define reg doesn't have init anymore (#11365) George HotzandGitHub 2025-07-24 19:15:49 -07:00
• 9da3f72495 identity store for DEFINE_REG (#11363) George HotzandGitHub 2025-07-24 16:41:29 -07:00
• cc795c6656 simplify keccak pad mask code (#11362) chenyuandGitHub 2025-07-24 19:24:10 -04:00
• 5bc73e644d noop continue id_store geohot 2025-07-24 16:20:53 -07:00
• 984a0edbc9 identity store for DEFINE_REG geohot 2025-07-24 16:13:43 -07:00
• 0b19e2dddd identity store for DEFINE_REG geohot 2025-07-24 16:04:58 -07:00
• dc68b7ef96 in_loop before2 geohot 2025-07-24 14:19:19 -07:00
• 9997f79c0a this is the right way to do this geohot 2025-07-24 14:09:43 -07:00
• 386bbf311c Ops.BEFORE finally fixes the DEFINE_REG hacks geohot 2025-07-24 14:02:45 -07:00
• c0c4bc9d7c use int32 for keccak reorder_indexes (#11360) chenyuandGitHub 2025-07-24 15:54:50 -04:00
• 0602b22086 kernel spec (#11359) George HotzandGitHub 2025-07-24 12:45:38 -07:00
• 519f1d13cc viz: generic stuff from gpu counters ui (#11358) qazalandGitHub 2025-07-24 20:29:24 +03:00
• 3b3de8df61 hcq: graphed copies (#11302) nimlgenandGitHub 2025-07-24 17:36:19 +03:00
• 3046ead6e8 jit: graph reports ei support (#11356) nimlgenandGitHub 2025-07-24 16:35:10 +03:00
• bf12041910 hcq: mapping of cpu to all hcq devices (#11354) nimlgenandGitHub 2025-07-24 12:52:38 +03:00
• 82e6de7fc6 more keccak reference tests (#11329) chenyuandGitHub 2025-07-23 22:06:39 -04:00
• 0eec0e3dc0 continue work on amd_uop_matmul continue_mat geohot 2025-07-23 18:49:19 -07:00
• b0dc97d1f7 write out kernel 3 in uops (#11352) George HotzandGitHub 2025-07-23 17:32:38 -07:00
• ece4604825 cleanups manual_kernel_3 geohot 2025-07-23 16:36:37 -07:00
• 0cc71fec59 bugfix to match speed geohot 2025-07-23 16:24:38 -07:00
• 1e56fb559a gemm passes spec geohot 2025-07-23 15:50:01 -07:00
• a87bc6dacc matmul is correct geohot 2025-07-23 15:35:57 -07:00
• 5b570196e4 support DEV= to specify device (#11351) chenyuandGitHub 2025-07-23 17:40:55 -04:00
• 68db0e2338 write out kernel 3 in uops geohot 2025-07-23 14:37:48 -07:00
• 76a2ddbd78 Move remote tests out of onnx (#11310) 3 people 2025-07-24 01:25:55 +05:00
• 7f0a41df4d move optional out of devectorize [pr] (#11350) George HotzandGitHub 2025-07-23 11:26:05 -07:00
• 0f374e10d2 cpu: use mmap for allocations (#11349) nimlgenandGitHub 2025-07-23 20:30:18 +03:00
• ae07a93814 simple block barrier (#11341) George HotzandGitHub 2025-07-23 10:14:11 -07:00
• 86e7504111 mypy check extra/onnx.py (#11348) chenyuandGitHub 2025-07-23 12:42:59 -04:00
• 960da9319d Remove StrEnum in onnx for python 3.10 (#11345) chenyuandGitHub 2025-07-23 11:52:25 -04:00
• 478a355325 gate PRINT_MATCH_STATS behind graph_rewrite tracking (#11344) qazalandGitHub 2025-07-23 16:32:43 +03:00
• ca09c180dc cpu: remove del spam (#11343) nimlgenandGitHub 2025-07-23 12:02:37 +03:00
• 304eb9cecb allocate less memory in am tests (#11342) nimlgenandGitHub 2025-07-23 11:11:26 +03:00
• e14b4fefa5 ranges on store (#11334) George HotzandGitHub 2025-07-22 21:00:50 -07:00
• c65b5aab62 small things from endrange (#11339) George HotzandGitHub 2025-07-22 19:45:37 -07:00
• 50be175f88 store endrange geohot 2025-07-22 19:28:39 -07:00
• 33d80db64c this can happen later geohot 2025-07-22 19:12:37 -07:00
• 28b6cff043 broken geohot 2025-07-22 19:05:11 -07:00
• 23ee2769bd gate store geohot 2025-07-22 19:03:06 -07:00
• 5055668d1e oops, forgot that geohot 2025-07-22 18:54:14 -07:00
• 8305a5804c just the range thing geohot 2025-07-22 18:52:26 -07:00
• 911bb4ff44 Merge branch 'master' into endrange George HotzandGitHub 2025-07-22 18:43:22 -07:00