5 people 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]>
Co-authored-by: George Hotz <[email protected]>
Co-authored-by: chenyu <[email protected]>

2025-04-08 21:54:04 -04:00

..

autogen

RDNA 3.5 support (#9627)

2025-03-31 01:15:20 +08:00

graph

fixes from the dsp branch + 12500 lines (#9683)

2025-04-02 13:07:17 +08:00

support

am: rdna 4 support (#9621)

2025-03-29 23:16:27 +07:00

__init__.py

pylint runtime/ and shape/ (#5044)

2024-06-18 19:48:18 -04:00

ops_amd.py

fix RDNA2 support (#9700)

2025-04-05 18:42:47 +03:00

ops_cloud.py

few type cleanups [pr] (#8347)

2024-12-20 01:56:01 -05:00

ops_cpu.py

CLANG -> CPU (#9189)

2025-02-20 18:03:09 -05:00

ops_cuda.py

bump assembly debug to 7 (#9662)

2025-04-01 11:51:33 +08:00

ops_disk.py

hotfix: load >2 GiB from disk on macOS (#9361)

2025-03-07 14:51:58 +08:00

ops_dsp.py

dsp: make relro sections contiguous (#9701)

2025-04-02 18:02:16 +07:00

ops_gpu.py

replace Tuple with tuple [pr] (#8344)

2024-12-19 21:27:56 -08:00

ops_hip.py

AMDCompiler refactor (no_comgr prereq) (#9497)

2025-03-20 09:44:07 +08:00

ops_llvm.py

refactor llvm compiler (#9403)

2025-03-18 00:13:49 +08:00

ops_metal.py

use functools.cache instead of lru_cache(None) [pr] (#9714)

2025-04-03 11:47:13 +08:00

ops_npy.py

add npy as_buffer (#9309)

2025-03-01 12:34:29 +03:00

ops_null.py

use Ops.REDUCE (#9721)

2025-04-04 10:14:34 +08:00

ops_nv.py

HCQ: Increment timeline signal before submitting (#9550)

2025-03-23 18:30:38 +07:00

ops_python.py

FP8 support on NVIDIA (#8631)

2025-04-08 21:54:04 -04:00

ops_qcom.py

hcq: dyn alloc signals (#9238)

2025-02-25 17:22:24 +03:00

ops_webgpu.py

types for WebGPU runtime (#9791)

2025-04-08 22:52:17 +03:00