George Hotz
affd83961c
small changes from define_reg ( #11327 )
...
* small changes from define_reg
* fix webgpu
2025-07-22 11:11:48 -07:00
George Hotz
3b674df34b
generic changes from define_reg_2 ( #11315 )
...
* generic changes from define_reg_2
* fix for ptx
* ugh, that one
2025-07-21 15:14:06 -07:00
chenyu
54924f9969
type remove Union and Optional [pr] ( #11283 )
...
use `|` for consistency
2025-07-19 14:05:52 -04:00
chenyu
ec3efd2919
move upcast before reduce ( #11250 )
...
* move upcast before reduce
upcast goes to end of global+local+upcast
* r_196_32_4_24_8
2025-07-18 14:42:15 -04:00
chenyu
522dc72f08
remove Kernel.local_dims [pr] ( #11268 )
...
* remove Kernel.local_dims [pr]
also not needed
* fix test_matvec
2025-07-16 17:46:19 -04:00
chenyu
c8e5c4d7c3
insert_before -> insert_at [pr] ( #11257 )
...
more precise
2025-07-15 17:44:34 -04:00
chenyu
b6662096cb
remove more first_reduce [pr] ( #11239 )
2025-07-14 19:13:44 -04:00
chenyu
eb8e17ef59
remove most of the first_upcast [pr] ( #11238 )
2025-07-14 16:54:24 -04:00
chenyu
674dc28505
remove Kernel.full_unupcasted_shape [pr] ( #11215 )
...
decomp to shape_len and first_upcast to get the last upcast-able dim
2025-07-13 13:56:23 -04:00
chenyu
2b48b961be
fix a few broken AMX tests ( #11204 )
2025-07-12 21:42:38 -04:00
chenyu
a0438012af
remove Kernel.get_program [pr] ( #11203 )
2025-07-12 20:50:29 -04:00
chenyu
6283d50224
DEPRECATED_linearize -> to_program [pr] ( #11198 )
2025-07-12 13:46:20 -04:00
George Hotz
2893feb9f6
cleanups for kernel.py ( #11143 )
...
* cleanups for kernel.py
* fixups
2025-07-08 18:10:25 -07:00
George Hotz
359bed74f8
axis type tracking [pr] ( #11137 )
...
* axis type tracking [pr]
* keep update_info
* keep legacy colors
* update tests to apply_opt
2025-07-08 14:16:25 -07:00
George Hotz
0597735f28
remove TC=3 not porting this ( #11045 )
2025-06-30 15:12:49 -07:00
chenyu
126fcf4129
clean up AMD_LLVM in tests ( #11021 )
2025-06-28 22:45:47 -04:00
George Hotz
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
5a1911b7c4
apply the global dims late ( #11002 )
...
* apply the global dims late [pr]
* late gpudims
* tests passing
* remove the random local_dims inc
* simpler
2025-06-27 09:54:34 -07:00
George Hotz
b4eb876d5a
kernel.py no longer permutes reduce axis [pr] ( #10968 )
...
* kernel.py no longer permutes reduce axis [pr]
* delete tests that handcode uops
* regen of sops is broken...
* put import back
* just remove that
* disable those tests
2025-06-26 17:44:58 -07:00
Ignacio Sica
579194f523
remove some linearize calls from tests 2 [pr] ( #10992 )
...
* refactor count_float4 to take uops as input instead of kernel
* remove some calls to linearize in test_linearizer
* remove some more calls
* remove one more call
2025-06-26 18:22:27 -03:00
Ignacio Sica
21f1c4cc09
remove some linearize calls from tests [pr] ( #10978 )
...
* remove some linearize calls from tests
speed_compare_cuda_ptx
test_uop_spec
test_linearizer
test_uops
test_winograd
* more clear assert message
2025-06-25 12:37:17 -07:00
Ignacio Sica
98d2cde293
revert tc_group feature ( #10971 )
2025-06-24 20:58:13 -07:00
George Hotz
8a65720528
hotfix: disable test_tensor_core_opts_group test on real metal
2025-06-24 15:21:33 -07:00
George Hotz
8743ca40e2
force reduce to be in axis order ( #10837 )
...
* force reduce to be in axis order
* disable rule causing loop
* disable that rule
* no ra there
* only move non reduce
* fix tests
2025-06-24 13:00:16 -07:00
Ignacio Sica
956a8391a5
minor cleanup on test_tensor_core_opts tests ( #10924 )
...
* minor cleanup on test_tensor_core_opts tests
Tests now notify when skipped
Before, they silently skipped if backend didn't had half precision and
accumulation
Also cleaned up atol and rtol setup
* refactor test_tensor_core_opts_group
---------
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com >
2025-06-23 16:30:21 -07:00
Ignacio Sica
b8d09a1dae
tc with group/grouptop ( #10903 )
2025-06-23 09:58:41 -07:00
George Hotz
92678e59ee
move kernel to opt ( #10899 )
2025-06-20 15:22:28 -07:00
George Hotz
32e9949052
rename lazydata to uop ( #10698 )
2025-06-08 08:42:22 -07:00
qazal
5b59728c75
refactor LOAD(DEFINE_GLOBAL, VIEW) in kernels to LOAD(VIEW(DEFINE_GLOBAL)) ( #10541 )
...
* changes to core tinygrad
* fixups pt1
TC=3
docs/abstractions2.py
IMAGE=2
test_quantize_dsp
test_schedule
* more tests
* green now
* images stay images
2025-05-30 14:27:58 +03:00
qazal
bbf05110a2
use kernelize in TestLinearizer.test_indexing_multireduce [pr] ( #10571 )
2025-05-30 11:27:09 +03:00
qazal
9169dcfb49
do not create kernels with more inputs than the backend allows ( #10510 )
...
* work
* no itertools + top down pass
* clean viz
* python can do that
* webgpu
* gbarrier of gbarrier is gbarrier
* device can be tuple
* bug in toposort
* failing test for gated toposort
* contiguous of gbarrier is gbarrier
* check for binops
* Revert "check for binops"
This reverts commit 53e3cdf720
2025-05-26 18:02:03 +03:00
George Hotz
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
Ignacio Sica
8f79492c75
fix test_tensor_cores_codegen for ptx renderer ( #10119 )
2025-05-01 21:52:36 -03:00
Ignacio Sica
bf5fb97498
fix AMD_LLVM bf16 tc for gfx1100 ( #10102 )
...
* fix amd_llvm bf16 tc
* cleanup pattern
2025-04-30 20:06:38 -03:00
Ignacio Sica
bda116d773
fix use_tensor_cores propagation ( #10048 )
...
* propagate use_tensor_cores
* add use_tensor_core to arg in test and search
* bugfix
* get TC val from ContextVar in search
* revert minor space change
* add tc emulation test to ci and benchmark
* revert
* revert whitespace change
* remove test for ptx
* add comment and remove llvm test run
2025-04-28 19:30:50 -03:00
George Hotz
4c242b0483
hotfix: tests all pass on metal local
2025-04-28 12:09:00 -04:00
qazal
d13c100981
don't sort dims in verify_sink_dims [pr] ( #10059 )
...
* don't sort dims in verify_sink_dims [pr]
* 1 can exist with n
* put process_replay warn last
* assert shape is the same
* bring that back
2025-04-26 23:24:30 +08:00
Ignacio Sica
76a86735c0
hotfix amd bf16 is supported case ( #10039 )
...
* hotfix amd and amd_llvm
* bf16 not supported in ci
* hotfix amd_llvm is not a device
* remove default
* dont gate on ci and amd_llvm
* minor cleanup
* skip bf16 tc test for amd_llvm
2025-04-24 21:29:27 -03:00
Ignacio Sica
b4f823acbe
fix helper_tc_allclose ( #9606 )
...
* fix helper_tc_allclose
* cleanup
* hotfix
* cleanup
* cleanup
* check real buffer and add cast for bf16
* cleanup
* fix padded for ops_python
* avoid assert on amd emulated tc
* swap dimensions
* revert, should have nothing to do with padded
* revert fix, should not go in this pr
* remove skip
2025-04-24 18:36:40 -03:00
Ignacio Sica
51ca19d061
set test_tensor_cores_padded_amd to expectedFailure ( #10036 )
...
* init
* add expected failure to correctly track progres
* hotfix
* skip for amd_llvm as well
* add skip
* add pr number
* move comment to amd test
* change reason
2025-04-24 17:11:40 -03:00
Ignacio Sica
373ca59b7f
use is_dtype_supported to check dtype support in tc tests ( #10035 )
2025-04-24 14:59:14 -03:00
George Hotz
2ed3acd767
toposort is a function [pr] ( #10004 )
2025-04-23 16:25:03 +01:00
chenyu
6c30948df6
hand_coded_optimizations returns list[Opt] [pr] ( #9938 )
...
new api looks like `k.apply_opts(hand_coded_optimizations(k))`
2025-04-19 20:26:59 -04:00
Ignacio Sica
023b1c28a2
test_tensor_cores_padded refactor (#9724 )
...
* set pad t 3 for amd padded tc test
* change pad for amd regardless CI
* test tc padded uops and correctness separately
* add test_tensor_cores_padded_uops test to ci
* remove redundant chack for amd device
* cleanup
2025-04-18 17:05:54 -03:00
George Hotz
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
chenyu
f5256e0020
Kernel.apply_opts [pr] ( #9917 )
...
* Kernel.apply_opts [pr]
updated all `for opt in`. also updated a few test_liinearizer tests to not implcitly depend on hand_coded_optimization
* not you yet
2025-04-17 08:00:56 -04:00
chenyu
8c6299bced
move hand_coded_optimizations to heuristic.py [pr] ( #9844 )
...
* move hand_coded_optimizations to heuristic.py [pr]
also folded all long lines
* make a copy and rename self -> k
* fix test
2025-04-10 23:40:16 -04:00
George Hotz
78caf55154
Revert "FP8 support on NVIDIA ( #8631 )"
...
This reverts commit 2c8e4ea865
2025-04-09 12:27:41 +08:00
pkotzbach
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 <pawkotz@gmail.com >
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com >
Co-authored-by: chenyu <chenyu@fastmail.com >
2025-04-08 21:54:04 -04:00
Ignacio Sica
58785181a8
AMD bf16xf32 TC ( #9717 )
...
* dont test bf16 for emulated amd tc
* skip bf16 tc test in ci
* skip bf16 for AMD in test_tensor_cores_codegen
* add simple bf16 gemm test to benchmark
2025-04-07 11:41:04 +08:00
