George Hotz and GitHub
dd2de4f838
rename all DEFINE_GLOBAL to PARAM ( #14511 )
2026-02-03 15:09:38 +08:00
nimlgen and GitHub
26220a472e
no core_id ( #14265 )
...
* no core_id
* kwargs
* est
* linters
* ugh
* revert this
* deps
* glb
* should work?
* nn
* line
* fx
* ym
* z
* d
* um?
* revert
* this one?
* first half
* um p2
* all?
* um
* cleaner
* um
2026-01-23 21:30:12 +03:00
qazal and GitHub
b1c5a242b7
Revert "move is_dtype_supported logic to renderer ( #14188 )" ( #14237 )
...
This reverts commit 161fee9a48
2026-01-20 12:19:14 +09:00
b65dc9fd8e
refactor: use generic type for ContextVar [pr] ( #13998 )
...
* use generic type for context var
removes ops_python string cast thing, allows for handling of other string vars like `_CC`
* update Context.old_context type
---------
Co-authored-by: chenyu <[email protected]
2026-01-19 13:37:54 -05:00
sirhcm and GitHub
161fee9a48
move is_dtype_supported logic to renderer ( #14188 )
...
* move is_dtype_supported logic to renderer
* fix CPU_COUNT
* mypy happy
* early import libclang too with llvm
* run with debug
* skip autogen tests if MTLCompiler or llvm is loaded
* run autogen tests separately in CI
* lint
2026-01-18 22:37:04 -05:00
chenyu and GitHub
35c9701df0
update outdated tests and comments ( #14090 )
2026-01-10 01:00:48 -05:00
chenyu and GitHub
3caa1e2c98
fix cast HALF with PYTHON backend ( #14058 )
2026-01-07 16:52:05 -05:00
nimlgen and GitHub
77a76d1b13
device: respect compiler ContextVars ( #13523 )
...
* device: envvars for cc
* fix
* fix
* x
* um
* fix
* remote
* em
* cleanup
* typing
* fix
* debug
* lvp?
* ugh
* singl
* rm
* lol
* fix
* ?
* this?
* why?
* rev
* mod test
* l
2025-12-02 14:42:04 +03:00
George Hotz and GitHub
5e01cc299b
zero len ranges fail ( #12974 )
...
* zero len ranges fail
* fix Python backend
* fix llvm
* fix ptx
* yolo fix nir
* this works...
* always store...
* always store...
* Revert "always store..."
This reverts commit 0816cf344d
2025-10-28 22:49:55 +08:00
George Hotz and GitHub
c0c24d3a70
cleanup wmma ( #12927 )
...
* cleanup wmma
* fix test_ops failures on android
2025-10-26 18:26:47 +08:00
George Hotz and GitHub
6415e3e8a7
use Ops.GROUP instead of Ops.NOOP for merging stores ( #12912 )
...
* use Ops.GROUP instead of Ops.NOOP for merging stores
* fs noop
2025-10-25 12:26:12 +08:00
George Hotz and GitHub
e85cee0aad
flip Ops.END srcs ( #12882 )
...
* flip Ops.END srcs
* backward
* late end split
2025-10-23 12:47:50 +08:00
b1tg and GitHub
81108f91ee
amd tc: 16x16x32 ( #12874 )
...
* amd tc: 16x16x32
* test
* clean, test amd_cdna4
2025-10-22 13:48:01 -04:00
b1tg and GitHub
60d7e232f2
cuda fp8 ( #12782 )
...
* cuda fp8
* tensor core
* tc test
* clean
* clean pm
2025-10-21 15:05:25 -04:00
George Hotz and GitHub
a71a41f6d1
rename Ops.ENDRANGE -> Ops.END ( #12824 )
2025-10-21 11:32:18 +08:00
George Hotz and GitHub
b8a9cce783
replace NOOP with AFTER in reg init ( #12804 )
...
* after op
* fix tests
* replace NOOP with AFTER in reg init
* closer
* or_after there
* fix device
* fix all renderers
* better spec for after
2025-10-20 15:34:32 +08:00
54c15d74a4
python float8 support ( #11960 )
...
* basic support
* alu
* nan in exec_alu
* rand_for_dtype
* inf + 0.0
* finfo
* revert rand_for_dtype
* clean
* truncate fp8s inf
* spec ok
* float_to_fp8 nan/inf
* least_upper_dtype
* clean up
---------
Co-authored-by: b1tg <[email protected]
2025-09-18 09:17:09 -04:00
nimlgen and GitHub
fb96394ff5
auto-select available compilers ( #12094 )
...
* device: auto select compilers
* fix
* metal+opencl
* nv/cuda
* test without ptx
* ptx
* fix tests
* fix
* fix test
* rename
* test + cleaner
* xx
* ops
* better test
* win?
* um?
* types
* debug
* win??
* sep rung
* wtf?
* debug
* skip win
* revert this
* types
2025-09-10 19:52:01 +03:00
George Hotz and GitHub
9dee724fc4
make EMULATE a context var ( #12002 )
...
* make EMULATE a context var
* fix test amx
2025-09-04 11:15:43 -07:00
Sieds Lykles and GitHub
572a3c15c6
Move Ops.SPECIAL arg to src ( #11918 )
...
* initial moving bound to src
* arg to src
* remove import
* fixup linearizer
* arg to src
* fix test_uop_graph
* fix more tests
* fix python renderer
* get const value from const uop
* ssimplify uop estimates
* fix webgpu locals
* fix old test
* gate Ops.SPECIAL in linearizer
* use ssimplify() for local/global_size
* remove toposort gate_parents_instead_of_self
* fix rendering in comment
* cleanup
* rename and add comments
* add BottomUpGate with test
2025-09-04 09:31:44 +02:00
75d380a77c
fix transcendentals in python renderer ( #11932 )
...
* fix transcendentals in python renderer
* add test
---------
Co-authored-by: b1tg <[email protected]
2025-08-31 09:37:17 -04:00
b2cc06218a
python bfloat16 ( #11912 )
...
* python bf16
* _to_torch_storage_type
---------
Co-authored-by: b1tg <[email protected]
2025-08-29 15:18:02 -04:00
chenyu and GitHub
5d6963c968
RuntimeError for unsupported dtype in PYTHON ( #11686 )
2025-08-15 13:59:27 -04:00
George Hotz and GitHub
82be8abfd2
move opt under codegen ( #11569 )
2025-08-07 14:19:17 -07:00
8dfcdb123d
less wmma args ( #11385 )
...
* less wmma args
* scalar
* ops_python
* mypy
* lint
* dedup
* helper wmma_args
---------
Co-authored-by: b1tg <[email protected] [email protected]
2025-07-26 21:24:05 -07:00
George Hotz and GitHub
466ab5a3f2
store/load not pass through index ( #11381 )
...
* noop
* fix noop
* store cat is NOOP
* store dtype is void
* stores aren't passed through anymore
* meh, skip those for ptx
* correct ptx skip
* hl runs
2025-07-25 21:01:47 -07:00
George Hotz and GitHub
490a93902c
define reg doesn't have init anymore ( #11365 )
...
* define reg doesn't have init anymore
* remove that
* no special logic for dr
* fix amd uop matmul
2025-07-24 19:15:49 -07:00
George Hotz and GitHub
9da3f72495
identity store for DEFINE_REG ( #11363 )
...
* identity store for DEFINE_REG
* identity store for DEFINE_REG
* noop continue
2025-07-24 16:41:29 -07:00
George Hotz and GitHub
c65b5aab62
small things from endrange ( #11339 )
...
* small things from endrange
* store
2025-07-22 19:45:37 -07:00
George Hotz and GitHub
09431d4ad1
make DEFINE_REG behave like the others ( #11273 )
...
* simpler define reg
* cast
* PTRCAT define_acc
* cleanups
* fix uops stats
* fix linearizer tests
* llvm
* define reg sets const
* define reg sets const
* no assign
* collapse that
* fix test_max_pool2d_bigger_stride_dilation
* use index, fix webgpu
* devec
* fix tests
* fix webgpu
* fix llvm
* threads for python
* fix ops_python
* only for reg
* acc_half is real now in the emulator
* fix llvm
* fix webgpu init
* fix wgpu test
* fix some tests
* fix ptx
* fix ptx bool acc
* cleanups
* broken, meh. will fix with ENDRANGE
* line count
2025-07-22 13:53:56 -07:00
chenyu and GitHub
54924f9969
type remove Union and Optional [pr] ( #11283 )
...
use `|` for consistency
2025-07-19 14:05:52 -04:00
George Hotz and GitHub
53ae153404
tc should be in opt ( #11148 )
...
* tc should be in opt [pr]
* fix import
2025-07-09 14:12:21 -07:00
George Hotz and GitHub
22305260e0
move tc to tc.py [pr] ( #11147 )
2025-07-09 10:55:56 -07:00
quortus and GitHub
17d85b9793
Refactor STORE implementation in ops_python ( #11060 )
2025-07-02 14:29:12 -07:00
George Hotz and GitHub
be53ef4f0a
rename DEFINE_ACC -> DEFINE_REG ( #11006 )
...
* rename DEFINE_ACC -> DEFINE_REG
* add CMPEQ to groupops
2025-06-27 11:09:25 -07:00
George Hotz and GitHub
411392dfb7
move files into uop dir ( #10399 )
...
* move files into uop dir [pr]
* tinygrad.uop is a thing
* fix uop docs, no pr
* fix viz
2025-05-18 11:38:28 -07:00
uuuvn and GitHub
7bc4864bc4
Make dev a property of Allocator ( #10286 )
...
* Make `dev` a property of `Allocator`
(this is a prereq refactor for #10285 )
At least `BufferXfer.copy` accesses it assuming it's always present,
currently most devices just add this property on their own repeating
the same code over and over again.
This is also a bit footguny, see `RemoteAllocator` that named this
property `device` instead of `dev`, i could obviously just change that
in one place but doing it globally seems like a better solution (and it
reduces code duplication too).
`MallocAllocator` is a bit special, but passing `None` works just fine.
* typing
* ignore type instead of cast
2025-05-13 17:01:01 -07:00
chenyu and GitHub
145e51247a
split CAST and BITCAST in PYTHON [pr] ( #10123 )
...
CAST only needs truncate and does not require dtype fmt. added bfloat16 tests can run locally
2025-04-30 23:27:35 -04:00
George Hotz and GitHub
c3ff308abb
range has only one src now [pr] ( #10100 )
...
* range has only one op now
* fix z3 checker
* ci fix
* needs shell
* try pip ensure update
* that ensurepip is useless
* upgrade pip before cache
* windows happy?
2025-04-29 10:31:05 -04:00
914d89fa0b
fix tensor cores for gfx1201 ( #9838 )
...
* fix tensor cores for gfx1201
* fix typo
* fix python wmma
* AMDLLVMRenderer with arch + AMDLLVM tensor_cores
* fix ci
* clean up
* more tensor cores for RDNA4
* fix half/half, bfloat16/float, bfloat16/bfloat16 for amd_llvm
---------
Co-authored-by: nimlgen <[email protected] [email protected]
2025-04-24 14:57:41 -04:00
George Hotz and GitHub
aa98aff4cd
don't use ops name, just keep sink ( #9922 )
...
* don't use ops name, just keep sink
* fix test
* endif sink
2025-04-18 08:59:18 +01:00
geohot
78caf55154
Revert "FP8 support on NVIDIA ( #8631 )"
...
This reverts commit 2c8e4ea865
2025-04-09 12:27:41 +08:00
2c8e4ea865
FP8 support on NVIDIA ( #8631 )
...
* squashed fp8 commits
* tensorcore start
* minor changes
* pre-commit
* pylint
* Delete fp8mul.cu
* clean
* small bugfix
* fix test_dtype
* fix test_dtype_alu
* add EMULATE_CUDA_SM89
* fix ci
* fix test_linearizer
* fix test_linearizer
* fix swizzle
* add debug to simple_matmul
* fixed swizzle
* python emulator
* refactor python emulator
* setup fix
* numpy setup
* ml_dtypes only in emulate_cuda_sm89
* fix pylint
* fix tests
* fix mypy
* fix mypy
* fix ruff
* done python emulator
* add acc type
* tests
* mypy
* clean code
* add cuda tensor core tests to CI
* minor fix
* clean test_dtype.py
* clean cstyle.py
* clean test_ops.py
* fix test
* fix test
* whitespaces
* pylint
* pylint
* amd?
* amd?
* amd
* reduce lines
* mockgpu remove
* fix
* ruff
* ruff
* fix mypy
* ruff
* test only for cuda
* fixed formatting
* small fixes
* small fix
* least_upper_dtype if fp8s not supported
* log and reciprocal are supported for fp8s
* ops python fixes
* dtypes.fp8s use
* e4m3 + e5m2 result dtype test
* truncate linter fix
---------
Co-authored-by: pkotzbach <[email protected] [email protected] [email protected]
2025-04-08 21:54:04 -04:00
Ignacio Sica and GitHub
bc91fffc5d
fix gated store with index in python backend ( #9703 )
...
* add default gate in index
* assert store
* add TestRendererFailures
- move test_gated_store_with_alu to new TestRenderFailures class for
tests that fail on multiple renderers
- add test_renderer_failures.py run on python CI
* add test for gated index in 2d
* test TestRenderFailures
2025-04-03 12:48:28 +08:00
qazal and GitHub
0b20f91ce7
remove move_mask from the devectorizer ( #9511 )
...
* remove move_mask from the devectorizer
* add (wrong) ptx
* reason
* enable index addition in PTX, we won't have the INDEX anyways
* space
2025-03-20 11:53:12 +08:00
Ignacio Sica and GitHub
5c56cac0a0
MI300 mfma support ( #9417 )
...
* add f16/f32 mfma support for MI300
- add 16x16 mfma shape support for f16 with f32 acc
- add ops_python mfma emulation
- add arch to AMDRenderer
* minor cleanup
* minor cleanup
* add mfma emulation task to ci
* add back todo
* hotfix: comment
* add tc=3 job to ci
2025-03-18 14:33:30 -03:00
chenyu and GitHub
2e7c2780a9
CLANG -> CPU ( #9189 )
2025-02-20 18:03:09 -05:00
George Hotz and GitHub
a4dab3ec3f
add name uop ( #9149 )
...
* add name uop, TODO: refactor renderer to use
* renderer uses name uop
* fix tests
* render
* ptx
2025-02-18 15:26:58 +08:00
Ali Ladjevardi and GitHub
6e523e4d17
Remove size arg from DEFINE_LOCAL [pr] ( #8845 )
...
* remove size arg form DEFINE_LOCAL
* make mypy happy
* whitespace
* dont change code in extra
* revert to temp1 to pass pr
2025-02-02 19:47:32 +08:00
d2234e308a
tf32 tc for nv and ptx ( #8635 )
...
Co-authored-by: George Hotz <[email protected]
2025-01-17 17:43:57 -08:00
sirhcmandGitHub
geohot