GitLab

* Rebased to main
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixed the year to 2026
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Added compilation guards
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Added BWD pass
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Added dbias and dact tests. Refactoring.
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Added grouped MXFP8 DACT and ACT API and tests
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixed a typo
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixes per the review
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* More fixes from the review
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixes per the review
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Relaxed requirement for last dim from mod128 to mod32
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fix
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Added alignment checks when tensor descriptors are modified
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



---------
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: vthumbe1503 <vthumbe@nvidia.com>

* version change
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* ifx
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

---------
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* code drop
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Add FP8 scale support and fix alignment for grouped GEMM

- Add FP8 scale_inv pointer handling in nvte_grouped_gemm for proper FP8 GEMM
- Fix random padding in tests to ensure 16-byte alignment for all dtypes
- Reorder GroupedGemmSetupWorkspace members for natural alignment
- Remove debug prints
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Grouped GEMM: code cleanup and NULL C support

- Remove unused alignment parameter from GroupedGemmSetupWorkspace::from_buffers
- Simplify select_grouped_operand by removing dead code branches
- Add GroupedOperandSelection.tensor field to avoid passing tensor separately
- Extract set_fp8_scale_pointers and init_matrix_layouts helpers
- Add safety check for FP8 on Hopper column-wise fallback
- Support NULL C tensor when beta=0 (uses D as placeholder)
- Remove unused get_scale_inv() from test
- Add use_null_c test parameter and test case
- Fix documentation: alpha/beta are single element tensors only
Signed-off-by: Piotr Gadzinski <pgadzinski@nvidia.com>
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Grouped GEMM: per-matrix alpha/beta support

- Change alpha/beta from single values to per-matrix arrays
- Validate alpha/beta have exactly num_tensors elements
- Update kernel to index alpha_ptr[idx] and beta_ptr[idx]
- Move alpha/beta validation to validate_grouped_gemm_inputs
- Update tests to use per-matrix alpha/beta arrays
- Update documentation
Signed-off-by: Piotr Gadzinski <pgadzinski@nvidia.com>
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fix alpha/beta numel - use SimpleTensor::numel()
Signed-off-by: Piotr Gadzinski <pgadzinski@nvidia.com>
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* Refactor: move grouped GEMM to separate file and cleanup API
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* Require Blackwell (SM100) and cuBLAS 13.1+ for grouped GEMM
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fixes
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* fixes
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Update transformer_engine/common/gemm/config.h
Co-authored-by: Przemyslaw Tredak <ptrendx@gmail.com>
Signed-off-by: Paweł Gadziński <62263673+pggPL@users.noreply.github.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* changed
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* suggestions
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* fix
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* refactored hopper tensor selection
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



---------
Signed-off-by: Pawel Gadzinski <pgadzinski@nvidia.com>
Signed-off-by: Piotr Gadzinski <pgadzinski@nvidia.com>
Signed-off-by: Paweł Gadziński <62263673+pggPL@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Przemyslaw Tredak <ptrendx@gmail.com>

* Disabled the tuned NVFP4 kernels
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Disabled fast math in cpp tests
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

---------
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Implemented persistent nvfp4 kernel
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fix FP4 guard in ptx
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fix
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fix in ptx. reduxf32 guard
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fix
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixes per PR review
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixes per PR review. Added parameter to turn off the persistency
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Modified reference CPU implementation in C++ unit tests to match GPU (numerical truncation). Tightened the numerical tolerance
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Disabled persistency by default, as non-persistent kernel is more performant when inputs are large
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Use the tuned kernel also for the rowwise only quantization
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixed typo
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Addressed comments from the PR review
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Resolved conflicts
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Macros renaming
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

---------
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Add general C API for setting tensor params
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Implement general accessors for NVTETensor
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Refactor tex swizzling to skip if scales are already swizzled
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Add checks for non-swizzled scales in MXFP8 and NVFP4 kernels
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Support pre-swizzled scales in MXFP8Tensor
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Add tex function to swizzle MXFP8 scales
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix bug in inplace swizzle function
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Tweak comments to use "compact/swizzled format"
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* MXFP8 quantize kernel with pre-swizzled scales
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Expose pre-swizzled scales in modules
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix bug in multi-swizzle
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Support MXFP8 gated activations with swizzled scales
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Add PyTorch infrastructure for pre-swizzled NVFP4 tensors
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Deprecate DSv3-specific quantization logic in C API
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Remove support for DSv3 compact data from quantizer
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Remove DSv3 compact data format from core lib
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix bug in FP8 all-gather
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix linter warnings
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Update JAX to use new swizzled scale API
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Review suggestion from @greptile-apps
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Review suggestions from @greptile-apps
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Update C++ swizzle test with swizzled scales API
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Return default tensor params when querying params for invalid NVTETensor
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Debug DSv3 FP8 test failures
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Debug Userbuffers test failures
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Make sure gated activations populate FP8 transpose if needed
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Review suggestions from @greptile-apps
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Disable pre-swizzling with debug quantizer
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Review suggestion from @greptile-apps
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix merge conflicts and review suggestions

Update copyright years. Tweak comments. Fix various complaints from @greptile-apps.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Use explicitly sized types in config accessors

Miscellaneous review suggestions from @ptrendx.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Make util header for function that compute swizzled scale index
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Apply suggestions from @greptile-apps
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Update expected error message in FP8 block-scaling test
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Review suggestion from @yaox12
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

Signed-off-by: wenjh <wenjh@sugon.com>

Signed-off-by: wenjh <wenjh@sugon.com>

Update copyright to include 2026
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

Fix amax computation using output_t data in normalization
Signed-off-by: Evgeny <etsykunov@nvidia.com>

* Removed activations from NVFP4 CPP tests. Removed CMake debugging flags
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Better wording
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

---------
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Update to_string(NVTEScalingMode) to include block scaling
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Add `nvte_swizzle_block_scaling_to_mxfp8_scaling_factors`
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Convert FP8 block scaling tensors to MXFP8 tensors on Blackwell and newer in GEMM
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Allow Blackwell and newer in Deepseek recipe compatbility check
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Allow data_rows % 4 != 0 in 1d kernel
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Load scaling factors in unswizzled order in 1d kernel
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Enforce use of power of two scaling
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Skip the FP8 block scaling exact GEMM test on Blackwell
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Skip further tests with pow_2_scales=False
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Initial implementation of tensor conversion for grouped gemm
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Skip non power of two scaling cpp unit tests
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Fix handling of all gather
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Apply suggestions from code review
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Use compute capability 10.0 for logic with Blackwell
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Apply suggestions from code review
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

---------
Signed-off-by: Jan Bielak <jbielak@nvidia.com>
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Add NVFP4 recipe
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Frank Sun <frsun@nvidia.com>
Co-authored-by: Oleg Goncharov <ogoncharov@nvidia.com>
Co-authored-by: Zhongbo Zhu <zhongboz@nvidia.com>
Co-authored-by: Evgeny Tsykunov <etsykunov@nvidia.com>
Co-authored-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Teddy Do <tdophung@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Add MathDx dependency to GitHub builds
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Suggestions from GitHub Copilot
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Move 2x shape logic from core to PyTorch
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Fix compilation errors with CUDA 12.1
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* SM 70 is not supported in CUDA 13
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Typo
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

* Revert "Move 2x shape logic from core to PyTorch"

This reverts commit f8b2a2d0111d9af690b43bb98ae448d9a430a185.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Added dequantize kernel for FP4
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fix linter warning
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Add NVFP4 support with fusible ops

Use logical tensor dims for PyTorch NVFP4 tensors. Temporarily add unfused dequantize impl. Fix bug where NVFP4 recipe was not configurable.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Fix logic for 2x shapes and move to PyTorch
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Fix CG test model config
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Debug NVFP4 tensor size function
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Proper handling of the RNG state
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* Test SR properly
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fix workspace size for GEMM heuristic.
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Fix compile error in C++ NVFP4 test

Some some numeric errors when blocks are all zero.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* fix distrbuted test problem shape
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* proper assert dim for low precision AG TP
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* clean up duplicated code in nvfp4_utils.cuh
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* lint
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* pylint: disable=unused-argument
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* `nvte_cublas_gemm_v2` to take alpha pointer (#12)

* make nvte_cublas_gemm_v2 to take alpha/beta pointers
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>

* users are expected to pass a valid C_tensor
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>

* typos
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>

* API to have const float* alpha
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>

* Minor tweaks

Support arbitrary beta scales. Increase workspace to be aligned to 128 bytes.
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Debug IMA with alpha pointer
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Support fused amax kernels with NVFP4 quantization
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Disable fused amax with cuDNN LayerNorm kernel
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Add NVFP4 cases to distributed tests for TE ops
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Change assert to NVTE_CHECK in the hadamard cast fusion
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* Fix compile error
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Use global thread IDs for Philox subsequences
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Add shape checks for NVFP4 cast kernels
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Do not fuse amax if cuDNN normalization is forced by envvar
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

---------
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Signed-off-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>
Signed-off-by: zhongboz <zhongboz@nvidia.com>
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>
Co-authored-by: Frank Sun <frsun@nvidia.com>
Co-authored-by: Oleg Goncharov <ogoncharov@nvidia.com>
Co-authored-by: Zhongbo Zhu <zhongboz@nvidia.com>
Co-authored-by: Evgeny Tsykunov <etsykunov@nvidia.com>
Co-authored-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Teddy Do <tdophung@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Co-authored-by: Przemek Tredak <ptredak@nvidia.com>
Co-authored-by: Phuong Nguyen <phuonguyen@nvidia.com>

Signed-off-by: wenjh <wenjh@sugon.com>

Signed-off-by: wenjh <wenjh@sugon.com>

* Add rmsnorm_bwd_add
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Add BackwardAddRMSNorm fused operation
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Try to optimize register usage in kernels
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Add separate BackwardAdd stage for the fused backward add
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

---------
Signed-off-by: Jan Bielak <jbielak@nvidia.com>

* Add basic kernel for swapping first two tensor dims
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Add NVRTC kernel for swapping first dims
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Add PyTorch extension for swap first dims kernel
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Tweak variable names
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Tune kernel
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Make sure writes are contiguous
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Fixed integer overflow when computing offsets
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



---------
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Fixed conflicts
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Minor code refactoring to avoid unnecessary checks
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixed typo
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixed dBias accumulation error due to initialization. Minor code refactoring
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Test case to reproduce the init error
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixed rowwise dbias error
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Changed ptx API
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Added a struct for two packed FP8 values
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Rolled back to scalar code for columnwise scaling due to its better performance
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Minor corrections
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Rebased on main
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixes per code review
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Removed constexpr in C++ test suite to build faster
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Computed activations are now numerically truncated to InputType before scaling. Improved test suite.
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Minor refactoring
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Minor refactoring
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Modified mismatches checks of MXFP8 to address FP8 numerics
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Implemented Jeremy's fixes to JAX test suite with an intermediate downcast
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Reduced the dims of the test tensors to improve CI runtime
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixed memory alignment issue. Compute dbias without downcast.
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixed misaligned memory issue also in gated kernels. Reduced size of MXFP8 gated tests
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



---------
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* enable cudnn norm tests
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>

* exclude tests on pre-Hopper
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>

---------
Signed-off-by: Phuong Nguyen <phuonguyen@nvidia.com>

Add env to chose blocklen of blockwise quantize.
Signed-off-by: wenjh <wenjh@sugon.com>

Fix pytest of blockwise error
Signed-off-by: wenjh <wenjh@sugon.com>

Resolve new api in  int8 gemm test
Signed-off-by: wenjh <wenjh@sugon.com>

Fix incorrect launch parm
Signed-off-by: wenjh <wenjh@sugon.com>

Fix 1D blockwise(64) acc error
Signed-off-by: wenjh <wenjh@sugon.com>

* [PyTorch|common] Implement unpadding kernel for FP8

1. Add multi-tensor unpadding kernel
2. Replace split+cat with unpadding kernel in Fp8Padding and Fp8Unpadding
3. Add unpadding with padding unit tests
Signed-off-by: xiaoxi-wangfj <690912414@qq.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* add license
Signed-off-by: Xin Yao <xiny@nvidia.com>

* Update padding.cu
Signed-off-by: Xin Yao <xiny@nvidia.com>

---------
Signed-off-by: xiaoxi-wangfj <690912414@qq.com>
Signed-off-by: Xin Yao <xiny@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Xin Yao <xiny@nvidia.com>

* Fix cppunittest test.sh for editable installs
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

* Update tests/cpp/CMakeLists.txt
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Fixes
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

---------
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

Signed-off-by: wenjh <wenjh@sugon.com>

* Added support of FP4 data type
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Refactoring to BitsNum in progress
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixed compilation errors. All C++ tests passed
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Fixed a typo
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Added FP4 guard to TMA tensor descriptor data type
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Fixed errors in JAX C++ extensions
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Removed dummy NVFP4 C++ test file
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Make pytorch changes
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Refactored the code per the review notes. Fixed JAX build error.
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* Removed unnecessary static casts
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>

* Typo fix
Signed-off-by: Oleg Goncharov <64355998+Oleg-Goncharov@users.noreply.github.com>

* Pass correct num bits to create_2D_tensor_map; fixes CI
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* inline funcs
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

---------
Signed-off-by: Oleg Goncharov <ogoncharov@nvidia.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Signed-off-by: Oleg Goncharov <64355998+Oleg-Goncharov@users.noreply.github.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

Signed-off-by: wenjh <wenjh@sugon.com>

Signed-off-by: wenjh <wenjh@sugon.com>

quantize_transpose_vector_blockwise function use lds exceeding 64kb when
input type is fp32. But max size of lds in dcu is 64kb, thus we use lds
as bfp16 for workaround.
Signed-off-by: wenjh <wenjh@sugon.com>

Use ocp fp8.
Workaround: test_cast_float8blockwise.cu link wrong std::max
Signed-off-by: wenjh <wenjh@sugon.com>

* Modify the test cases
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* Make the tests reproducible on different machines
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* Fixed the cache of the gamma_in_weight_dtype setting
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* Reinstate the tests
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



* More verbose code and comments
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>

---------
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Reapply "Allow NVTEShape to own data." (#1703)

This reverts commit 91405eb4

.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update code so that data is replaced by an array.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Specify unambiguous Tensor constructor in tests.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Fix assumption in test of 2D shape.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Remove row and col
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

---------
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

Signed-off-by: wenjh <wenjh@sugon.com>

Introduce nvte_memset to provide a fill kernel that is faster than cudaMemsetAsync for small sizes (#1716)

* nvte_memset fills single float value
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

* Support larger sizes than a single value and add tests
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

---------
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

* Add a flag to support computing zero-centered gamma in weight dtype or compute dtype for CuDNN
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

* Address comments
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

---------
Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

* Add GEMM logic for blockwise quantized tensors.

GEMM test cases included in pytorch integration.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update NVTE_BLOCK_SCALING for GEMM.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Gate feature on CUDA 12.9
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Gemm typo.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Remove unecessary type converter change.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Reflect epilogue availability and test supported epilogues.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* GEMM simplifications from recipe branch.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Format py code.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update GEMM DGelu tests to match support depending on output dtype.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Force pow2Scales in GEMM
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Add GEMM test to pytorch test suite.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Add copyright to GEMM test.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update import for GEMM test.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Add license.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update test gemm supported predicate.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Use sgemm like interfaces and naming.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Rewrite GEMM comment.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* MR Feedback.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Recipe setup for Linear modules.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Use 12.9 feature test.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Run against tensor dumps from internal library.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update FIXME to TODO with linked issue.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update full recompute feature to save recipe.

The recompute context uses the same recipe
and fp8 settings as the original fwd pass.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* MR Feedback. Avoid reusing quantizer objects.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update logic in module.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Format py.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update for PP bug.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update test numerics.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update force_power_of_2 scales in the recipe.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update usage method to satisfy upstream changes.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* fix subchannel recipe in distributed test with bf16 gather
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* Edit and cleanup BF16 gather code.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update test import.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* support columnwise only mode to 1D quantize kernel
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* Format and move enum
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Skip alloc.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* try async bf16 gather
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* Format python code.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Document and type code.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update pytorch lint errors.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Dont set high precision dtype.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Add test for sanity and CG; fix CG for sequential?
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

* Keep make_quantizers API stable

Update num_quantizers instead to pass cuda_graph tests.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Fix import name.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Rename recipe method.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Skip grouped linear sanity test.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Set usage before BF16 gather.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* refactor for nvte_quantize_v2
Signed-off-by: zhongboz <zhongboz@nvidia.com>

* Format code.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Cleanup nvte_quantize_v2
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Test fp32 scales.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Disable CUDA graph.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Simplify layernorm linear
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Cleanup layernorm linear.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* LayerNorm linear bwd gather logic.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Communication updates.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update transformer_engine/pytorch/ops/op.py

Apply MR comment change.
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Signed-off-by: kwyss-nvidia <kwyss@nvidia.com>

* Lint fix.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* MR feedback.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Enable cuda graph tests.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Reduce chance of spurious failure and reword.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Review suggestions from @timmoon10
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Update CPP tests.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update common.h
Signed-off-by: Xin Yao <yaox12@outlook.com>

* Update test_float8blockwisetensor.py
Signed-off-by: Xin Yao <yaox12@outlook.com>

---------
Signed-off-by: Keith Wyss <kwyss@nvidia.com>
Signed-off-by: zhongboz <zhongboz@nvidia.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Signed-off-by: kwyss-nvidia <kwyss@nvidia.com>
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Signed-off-by: Xin Yao <yaox12@outlook.com>
Co-authored-by: zhongboz <zhongboz@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Xin Yao <yaox12@outlook.com>

* Blockwise float8 quantizer and quantized tensor class.

The classes are configurable for 128x128 blocksize
and 1x128 blocksize via setting block_scaling_dim == 2,1 respectively.

Scale tensors are stored in a format emenable for matrix multiplication,
however the integration of matmul is deferred as a separate story.

Fusions of quantization and DBIAS or activation functions are not yet
implemented, and the dequantization is currently implemented in torch.

Tests for quantization are included in C++ and pytorch layers, with
exact comparison to reference quantizer behavior as well as an attempt
to hit interesting branches through the API such as tensor creation
in pytorch and CPP and dequantization of row and columnwise usage.

Two CUDA kernels for quantization are included, and are direct ports
of equivalents in the kitchen repository, where a subchannel recipe
has been used for end to end training.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Apply linting changes.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Alignment for 1D scaling for GEMM edge case.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* MR feedback.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Change API name.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Fix merge conflict with name change.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Use common tensor map API.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Change API to use two scaling mode enums.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Fix typo.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update some call sites.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Tests for torch tensor API surface.

Since the quantized tensor is a tensor
subclass, these tests exercise torch hooks.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Reuse scale calculation between quantizer refs.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Save memory by dropping reference to saved tensors.

Issues previously observed are solved.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Remove constexpr parameters from kernel.

Code size is reduced with fewer constexpr params.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Merge conflict from rebase.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Add shape implementations for block scaling.

nvte_shape was added upstream. Logic added
for block scaled fp8.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Move benchmark to te_playground
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Remove amax_epsilon and pow_2_scales from tensor.

Hardcodes the default values.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Lint changes.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Fixup MR changes that broke.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Safer ifdef in kernel.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Documentation prose.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Reuse compute_scale function from Current Scaling.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Bugfix on inf_value scale refactor.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Remove qopt calls from test.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update pytest list.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Add copyright to reference scale calc.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Use ptx.cuh functions instead of cde.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update shape logic with allocation and reuse shape.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Usage defaults MR feedback.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Copyright and header guard.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Updating torch dispatch code.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Fix exception type.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Use TypeInfo
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* MR feedback.
Signed-off-by: Keith Wyss <kwyss@nvidia.com>

* Update CS scale update test to use updated ref impl
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Update JAX scaling mode enum
Signed-off-by: Tim Moon <tmoon@nvidia.com>

* Skip tests on Lovelace
Signed-off-by: Tim Moon <tmoon@nvidia.com>

---------
Signed-off-by: Keith Wyss <kwyss@nvidia.com>
Signed-off-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Tim Moon <tmoon@nvidia.com>
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

Parallelize CPU reference implementation in tests
Signed-off-by: Evgeny Tsykunov <etsykunov@etsykunov-mlt.client.nvidia.com>
Co-authored-by: Evgeny Tsykunov <etsykunov@etsykunov-mlt.client.nvidia.com>
Co-authored-by: Oleg Goncharov <64355998+Oleg-Goncharov@users.noreply.github.com>