All issues

Search results

Require citing the faer paper when using the faer backend

Status: Open.

#1728 In tensor4all/tenferro-rs;

· shinaoka opened on Aug 25, 2026

Document downstream custom CUDA kernels with nvcc and NVRTC

Status: Open.

#1724 In tensor4all/tenferro-rs;

· shinaoka opened on Aug 24, 2026

[Performance]: CUDA SVD driver heuristic is ~2.9x slower than gesvd on RTX 3060

Status: Open.

#1720 In tensor4all/tenferro-rs;

· terasakisatoshi opened on Aug 22, 2026

[Bug]: #[allow(dead_code)] on flat_to_multi in tenferro-tensor/src/types.rs lacks adjacent // INVARIANT: marker (REPOSITORY_RULES.md:174-175)

Status: Open.

#1718 In tensor4all/tenferro-rs;

· tensor4all-ai-bot opened on Aug 21, 2026

[Docs]: AX guide (llms.txt) audit — einsum dialect restrictions are stated as neutral spec, not as warnings naming the failure mode

Status: Open.

#1715 In tensor4all/tenferro-rs;

· terasakisatoshi opened on Aug 20, 2026

[Bug]: CPU integer reduce_sum/reduce_prod "_wrapping" entry points are dead aliases — no explicit two's-complement wrapping in tenferro CPU code (REPOSITORY_RULES.md:102-111)

Status: Open.

#1707 In tensor4all/tenferro-rs;

· tensor4all-ai-bot opened on Aug 18, 2026

[Design] Re-evaluate CPU release-codegen boundaries after strided-kernel migration

Status: Open.

#1706 In tensor4all/tenferro-rs;

· shinaoka opened on Aug 18, 2026

[Performance]: eager leaf construction pays the ~5.6us backend session open (from_tensor_in ~12us for 8 elements)

Status: Open.

#1704 In tensor4all/tenferro-rs;

· lingrui96 opened on Aug 17, 2026

[Long-term design] Extract an IR-agnostic AD transform framework into a redesigned tidu-rs

Status: Open.

#1695 In tensor4all/tenferro-rs;

· shinaoka opened on Aug 15, 2026

Long-term: CPU linalg backend strategy — pure-Rust BLAS+LAPACK evaluation (faer / gemmkit / fused kernels / SME)

long-termLonger-horizon or deferred workLonger-horizon or deferred work

Status: Open.

#1660 In tensor4all/tenferro-rs;

· shinaoka opened on Aug 12, 2026

RFC: Align the contraction boundary with TAPP — conformance fixtures, a TAPP/TBLIS backend, and a TAPP C export

Status: Open.

#1658 In tensor4all/tenferro-rs;

· chenpeizhi opened on Aug 11, 2026

Investigate Mac CPU performance gaps in views, permutations, FFT, and small-shape workloads

enhancementNew feature or improvementNew feature or improvement

Status: Open.

#1628 In tensor4all/tenferro-rs;

· terasakisatoshi opened on Aug 6, 2026