NVIDIA/MatX v0.9.1

v0.9.1

Repository: NVIDIA/MatX

Tag: v0.9.1

Published: 2025-05-14T15:43:42Z

Prerelease: no

Release notes:

Sparse support + bugfixes

• New operators: argminmax, dense2sparse, sparse2dense, interp1, normalize, argsort
• Removed requirement for --relaxed-constexpr
• Added MatX NVTX domain
• Significantly improved speed of svd and inv
• Python integration sample
• Experimental sparse tensor support (SpMM and solver routines supported)
• Significantly reduced FFT memory usage

What's Changed

• Moving definition of CUB cache up by @cliffburdick in https://github.com/NVIDIA/MatX/pull/771
• Added documentation of memory types by @cliffburdick in https://github.com/NVIDIA/MatX/pull/770
• Cleaning up non-const operator() to avoid code duplication by @cliffburdick in https://github.com/NVIDIA/MatX/pull/769
• Switch to CUB/Thrust backend for cuda executor argmax by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/772
• Refactor cub argmax to generic cub reduce, use for argmin. Fixes #774. by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/776
• Change any() and all() to use CUB's reduce by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/777
• Add argminmax operator by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/778
• Fix matx::HostExecutor segfault with argmin/argmax by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/780
• Added new cusolverDnXsyevBatched API for batched eigen calls for CTK 12.6.2 and up by @cliffburdick in https://github.com/NVIDIA/MatX/pull/781
• cub.h CUDACC guards for custom ops by @nvjonwong in https://github.com/NVIDIA/MatX/pull/782
• Add example compiled with host compiler to catch regressions. by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/783
• Remove relaxed constexpr by @cliffburdick in https://github.com/NVIDIA/MatX/pull/775
• Cleanup versions.json so jq can parse it. by @alliepiper in https://github.com/NVIDIA/MatX/pull/785
• Allow rapids-cmake's version file to be overridden. by @alliepiper in https://github.com/NVIDIA/MatX/pull/786
• Update rapids-cmake (branch-24.12@03ec7ef) by @alliepiper in https://github.com/NVIDIA/MatX/pull/787
• Created MatX NVTX domain by @cliffburdick in https://github.com/NVIDIA/MatX/pull/784
• Update docs github action by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/789
• Update docs github action by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/790
• Work around compiler parser bug by @cliffburdick in https://github.com/NVIDIA/MatX/pull/791
• Updating developer documentation by @cliffburdick in https://github.com/NVIDIA/MatX/pull/793
• Modify concat op to enable concatenating float3. by @nvjonwong in https://github.com/NVIDIA/MatX/pull/792
• Fix rapids cmake by @alliepiper in https://github.com/NVIDIA/MatX/pull/799
• Switched to getRs instead of getRi for faster inverse by @cliffburdick in https://github.com/NVIDIA/MatX/pull/797
• Update CMakeLists.txt by @cliffburdick in https://github.com/NVIDIA/MatX/pull/801
• Support half precision R2C transforms by @cliffburdick in https://github.com/NVIDIA/MatX/pull/796
• Fix gcc13 erroneous warning by @cliffburdick in https://github.com/NVIDIA/MatX/pull/802
• fixed missing forwarding code for allocate by @aartbik in https://github.com/NVIDIA/MatX/pull/804
• Fix bug with eye, and also zero workspace before LU factorization by @cliffburdick in https://github.com/NVIDIA/MatX/pull/807
• Change shape_type for the remap op by @nvjonwong in https://github.com/NVIDIA/MatX/pull/806
• Faster batched SVD for small sizes by @cliffburdick in https://github.com/NVIDIA/MatX/pull/805
• Fixing broadcasting in all operator() by @cliffburdick in https://github.com/NVIDIA/MatX/pull/795
• Add a better error on memory allocation failure by @cliffburdick in https://github.com/NVIDIA/MatX/pull/808
• Fix solver interfaces to use executor in cache by @cliffburdick in https://github.com/NVIDIA/MatX/pull/809
• Python integration sample by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/812
• Fixes for clang17 errors/warnings by @cliffburdick in https://github.com/NVIDIA/MatX/pull/815
• Misc Cleanup by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/814
• frexp_fix by @cliffburdick in https://github.com/NVIDIA/MatX/pull/817
• Adding structures needed for sparse support by @cliffburdick in https://github.com/NVIDIA/MatX/pull/819
• fix missing newline at EOF (to avoid future diff issues) by @aartbik in https://github.com/NVIDIA/MatX/pull/822
• add size() to container storage by @aartbik in https://github.com/NVIDIA/MatX/pull/824
• minor edit for sparse (layout and proper swap def) by @aartbik in https://github.com/NVIDIA/MatX/pull/820
• add a to-string method for memory space by @aartbik in https://github.com/NVIDIA/MatX/pull/823
• Cleanup cmake usage when MatX is a dependent project by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/827
• Fixing warnings issues by clang-19, both host and device by @cliffburdick in https://github.com/NVIDIA/MatX/pull/825
• Update build_docs actions to newest. Add CI_RUN_DATETIME in version.rst by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/829
• introduce a versatile sparse tensor type to MatX (experimental) by @aartbik in https://github.com/NVIDIA/MatX/pull/821
• Add initial tiff support by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/831
• Make dim2lvl translation for printing more in the style of MatX by @aartbik in https://github.com/NVIDIA/MatX/pull/832
• Expose tensor format (and lvl specs) to sparse tensor data by @aartbik in https://github.com/NVIDIA/MatX/pull/833
• Add cross product operator by @mfzmullen in https://github.com/NVIDIA/MatX/pull/818
• remove LVL depth restriction with constexpr templating by @aartbik in https://github.com/NVIDIA/MatX/pull/834
• Guard all DIM/LVL recursion against completely empty format by @aartbik in https://github.com/NVIDIA/MatX/pull/835
• Adjust half-type threshold for cross product unit tests by @mfzmullen in https://github.com/NVIDIA/MatX/pull/838
• Added fp32 version of normcdf by @cliffburdick in https://github.com/NVIDIA/MatX/pull/839
• Changing black scholes to float and improving performance by @cliffburdick in https://github.com/NVIDIA/MatX/pull/840
• Implement the () operator on sparse tensors by @aartbik in https://github.com/NVIDIA/MatX/pull/837
• Support operators into einsum interface by @cliffburdick in https://github.com/NVIDIA/MatX/pull/845
• Add print function with nonzero dim args by @tbensonatl in https://github.com/NVIDIA/MatX/pull/844

*…