NVIDIA/MatX v0.9.2

v0.9.2

Repository: NVIDIA/MatX

Tag: v0.9.2

Published: 2025-07-29T19:13:10Z

Prerelease: no

Release notes: *New operator:* interp

*Other Additions:*

• Improvements to sparse support including new batched tri-diagonal solver
• Automatic vectorization and ILP support
• DLPack updated to 1.1
• Many bug fixes

What's Changed

• Fix partial any/all reduction by @simonbyrne in https://github.com/NVIDIA/MatX/pull/959
• interp1: add support for higher dimensional sample points and values by @simonbyrne in https://github.com/NVIDIA/MatX/pull/963
• Introduce DIA and SkewDIA format by @aartbik in https://github.com/NVIDIA/MatX/pull/964
• Refactor MATX_CUDA_CHECK to prevent multiple evaluation by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/957
• Introduce DIA format factory method by @aartbik in https://github.com/NVIDIA/MatX/pull/965
• reformat sparse files with clang-format by @aartbik in https://github.com/NVIDIA/MatX/pull/966
• Implement DIA SpMV kernel by @aartbik in https://github.com/NVIDIA/MatX/pull/967
• Generalize SpMV from square to m x n DIA by @aartbik in https://github.com/NVIDIA/MatX/pull/969
• replace static_assert(false) with host-only THROW by @aartbik in https://github.com/NVIDIA/MatX/pull/968
• Generalize DIA to DIA-I and DIA-J by @aartbik in https://github.com/NVIDIA/MatX/pull/972
• Avoid name collision with cpu_set_t from sched.h by @tbensonatl in https://github.com/NVIDIA/MatX/pull/971
• Add axis argument to interp1. by @simonbyrne in https://github.com/NVIDIA/MatX/pull/970
• Add operator tests back by @cliffburdick in https://github.com/NVIDIA/MatX/pull/977
• clang-format on sparse tests by @aartbik in https://github.com/NVIDIA/MatX/pull/973
• Add SpMV test for DIA-I and DIA-J by @aartbik in https://github.com/NVIDIA/MatX/pull/974
• (re) enable all sparse tests by @aartbik in https://github.com/NVIDIA/MatX/pull/979
• Let X = solve(A, B) take X and B along rows by @aartbik in https://github.com/NVIDIA/MatX/pull/981
• Add tri-diagonal solve support by @aartbik in https://github.com/NVIDIA/MatX/pull/982
• update doc with latest DIA support by @aartbik in https://github.com/NVIDIA/MatX/pull/983
• minor sparse documentation refinement by @aartbik in https://github.com/NVIDIA/MatX/pull/984
• Updating Google Test by @cliffburdick in https://github.com/NVIDIA/MatX/pull/985
• Minor fix in UST level order for DIA by @aartbik in https://github.com/NVIDIA/MatX/pull/986
• Vectorization and ILP by @cliffburdick in https://github.com/NVIDIA/MatX/pull/980
• Fixing compile error with FFT conv by @cliffburdick in https://github.com/NVIDIA/MatX/pull/989
• Fixing another 12.9 compiler bug by @cliffburdick in https://github.com/NVIDIA/MatX/pull/991
• Removing unused parameter in lambda causing error on clang by @cliffburdick in https://github.com/NVIDIA/MatX/pull/992
• proper lvl2dim computation for add/sub by @aartbik in https://github.com/NVIDIA/MatX/pull/994
• add braces to if-then-else by @aartbik in https://github.com/NVIDIA/MatX/pull/997
• Avoid fmod become ambiguous once CCCL specializes it for extended floating point types by @miscco in https://github.com/NVIDIA/MatX/pull/996
• clang formatting by @aartbik in https://github.com/NVIDIA/MatX/pull/998
• implement *batched* tri-diagonal direct solve by @aartbik in https://github.com/NVIDIA/MatX/pull/999
• add streams to alloc/free in cusparse sequences by @aartbik in https://github.com/NVIDIA/MatX/pull/1001
• test for batched tri-diag direct solver by @aartbik in https://github.com/NVIDIA/MatX/pull/1000
• fix minor typos in comments by @aartbik in https://github.com/NVIDIA/MatX/pull/1002
• DLPack 1.1 update by @cliffburdick in https://github.com/NVIDIA/MatX/pull/1004
• Fix host compiler errors when using -Wall -Werror by @tmartin-gh in https://github.com/NVIDIA/MatX/pull/1006
• Fix ARM relocation trucation build errors by @dylan-eustice in https://github.com/NVIDIA/MatX/pull/1008
• Allocate pinned host memory instead of managed when managed isn't available by @cliffburdick in https://github.com/NVIDIA/MatX/pull/1010
• Added executor to cache by @cliffburdick in https://github.com/NVIDIA/MatX/pull/1009
• Remove template parameters in constructor by @cliffburdick in https://github.com/NVIDIA/MatX/pull/1012
• fix flipud for 1D tensors by @simonbyrne in https://github.com/NVIDIA/MatX/pull/1011
• Fix warnings in clang19 by @cliffburdick in https://github.com/NVIDIA/MatX/pull/1015
• Missing unit test syncs by @dylan-eustice in https://github.com/NVIDIA/MatX/pull/1013
• add convenience constructor for batched tri diag sparse tensor by @aartbik in https://github.com/NVIDIA/MatX/pull/1019
• Remove runtime checks on memory spaces by @aartbik in https://github.com/NVIDIA/MatX/pull/1018
• build each test file as a separate executable by @simonbyrne in https://github.com/NVIDIA/MatX/pull/1017
• use batched sparse solve for interp by @simonbyrne in https://github.com/NVIDIA/MatX/pull/1016

New Contributors

• @miscco made their first contribution in https://github.com/NVIDIA/MatX/pull/996
• @dylan-eustice made their first contribution in https://github.com/NVIDIA/MatX/pull/1008

Full Changelog: https://github.com/NVIDIA/MatX/compare/v0.9.1...v0.9.2