All pull requests

Pull requests

• Lower f32 max/min to native max.f32/min.f32

  cuda-oxide
  #1668·CanReader opened Oct 28, 2026Contributor·2 tasks done·

  ·
  1

• Refresh stale example lockfiles

  cuda-oxide
  #1667·CanReader opened Oct 28, 2026Contributor·

  ·
  1

• fix(cutile-rs): report a missing cutile::core registry entry in from_kernel

  #1666·veljjanoski opened Oct 28, 2026·7 tasks done·

  ·
  3

• docs(cutile): add windows setup

  #1664·SickleFire opened Oct 28, 2026·3 of 5 tasks·

  ·
  1

• fix(cutile-rs): instantiate pointer-bound type parameters as pointers

  #1657·veljjanoski opened Oct 26, 2026·6 tasks done·

  ·
  1

  1

• fix(cuda-oxide): lower core::intrinsics::float_to_int_unchecked

  cuda-oxide
  #1654·veljjanoski opened Oct 26, 2026·5 tasks done·

  ·
  1

  1

• perf(cute): speed up the CuTe verifier's provenance analysis

  cuda-oxide
  #1648·juliankang4 opened Oct 26, 2026·4 of 5 tasks·

  ·
  1

  2

• docs(cutile): LoadTileLike zero-pads undersized inputs

  #1644·mangeshraut712 opened Oct 25, 2026·1 of 2 tasks·

  ·
  1

  2

• feat(cargo-oxide): opt-in CUDA_OXIDE_LINKER=wild host linker

  cuda-oxide
  #1641·deepakshirkem opened Oct 25, 2026·

  ·
  1

• fix(cutile): run a toolkit-resolved tileiras against its own toolkit

  #1637·0z5a opened Oct 24, 2026·3 of 5 tasks·

  ·
  1

• fix(cutile): build the reference-cpu examples on aarch64

  #1636·0z5a opened Oct 24, 2026·3 of 4 tasks·

  ·
  1

• fix(cutile): pass the smoke-test feature to the benchmark smoke run

  #1634·0z5a opened Oct 24, 2026·2 of 3 tasks·

  ·
  1

• test(cargo-oxide): keep an inherited CARGO_TARGET_DIR out of the configured-target test

  cuda-oxide
  #1633·0z5a opened Oct 24, 2026·1 of 2 tasks·

  ·
  1

• feat(cute-rs): accept sm_103a (B300) as a validated target

  cuda-oxide
  #1632·0z5a opened Oct 24, 2026·3 of 4 tasks·

  ·
  1

• fix(cuda-oxide): make check-error-example-status immune to SIGPIPE

  cuda-oxide
  #1630·0z5a opened Oct 24, 2026·1 of 5 tasks·

  ·
  1

• fix(cuda-oxide): gate the tcgen05 example on its module's target

  cuda-oxide
  #1629·0z5a opened Oct 24, 2026·2 of 5 tasks·

  ·
  1

  1

• fix(cutile-ir): reject invalid constant offsets

  #1627·DrunkTrader opened Oct 24, 2026·3 tasks done·

  ·
  2

• Support checked Block callbacks with uniform captures

  #1625·tpn opened Oct 23, 2026Contributor·

  ·
  1

• docs(cutile-rs): clarify GPU support below sm_80

  #1624·elibol opened Oct 23, 2026Collaborator·1 task done·

  ·
  1

• ci: run cargo deny on the cutile-rs workspace

  cicutile-rssupply-chain
  #1621·roivanov opened Oct 23, 2026Collaborator·3 tasks·

  ·
  1

  2

• release(v0.5.0): integrate shared host runtime and source-location fixes

  #1613·elibol opened Oct 22, 2026Collaborator·

  ·
  11

• fix(ci): check source references under docs

  cuda-oxide
  #1611·PavanaSakethaRam opened Oct 22, 2026·2 tasks done·

  ·
  1

  1

• ci(cutile-rs): complete monorepo cleanup

  #1610·PavanaSakethaRam opened Oct 22, 2026·

  ·
  1

  1

• Unify the async host runtime and remove cuda_async::simt

  #1604·elibol opened Oct 21, 2026Collaborator·

  ·

  · cuda oxide v0.4.0
  3

• refactor(cuda-core): share host handles and allocations across Tile and SIMT

  #1603·elibol opened Oct 21, 2026Collaborator·

  ·

  · cuda oxide v0.4.0
  2