Alternatives and similar repositories for torchair

Users that are interested in torchair are comparing it to the libraries listed below

• Ascend / triton-ascend
  Triton adapter for Ascend. Mirror of https://gitee.com/ascend/triton-ascend
  ☆107Updated this week
• feifeibear / LLMRoofline
  Compare different hardware platforms via the Roofline Model for LLM inference tasks.
  ☆120Updated last year
• tile-ai / TileOPs
  ☆84Updated last week
• DeepLink-org / DLSlime
  DLSlime: Flexible & Efficient Heterogeneous Transfer Toolkit
  ☆92Updated 2 weeks ago
• flashinfer-ai / cutlass-viz
  ☆65Updated 9 months ago
• InternLM / turbomind
  ☆96Updated 10 months ago
• AlibabaPAI / FLASHNN
  ☆105Updated last year
• infinigence / Semi-PD
  A prefill & decode disaggregated LLM serving framework with shared GPU memory and fine-grained compute isolation.
  ☆123Updated last month
• LeiWang1999 / AutoGPTQ.tvm
  GPTQ inference TVM kernel
  ☆40Updated last year
• thunlp / Seq1F1B
  Sequence-level 1F1B schedule for LLMs.
  ☆38Updated 5 months ago
• tlc-pack / libflash_attn
  Standalone Flash Attention v2 kernel without libtorch dependency
  ☆114Updated last year
• flagos-ai / libtriton_jit
  A Triton JIT runtime and ffi provider in C++
  ☆31Updated 2 weeks ago
• LeiWang1999 / Stream-k.tvm
  ☆20Updated last year
• inclusionAI / asystem-amem
  A NCCL extension library, designed to efficiently offload GPU memory allocated by the NCCL communication library.
  ☆91Updated last month
• KuangjuX / NVSHMEM-Tutorial
  NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer
  ☆161Updated 4 months ago
• Bruce-Lee-LY / decoding_attention
  Decoding Attention is specially optimized for MHA, MQA, GQA and MLA using CUDA core for the decoding stage of LLM inference.
  ☆46Updated 8 months ago
• ademeure / DeeperGEMM
  DeeperGEMM: crazy optimized version
  ☆73Updated 9 months ago
• KuangjuX / AttnLink
  An experimental communicating attention kernel based on DeepEP.
  ☆35Updated 6 months ago
• DeepLink-org / DLBlas
  DLBlas: clean and efficient kernels
  ☆33Updated this week
• infinigence / FlashOverlap
  A lightweight design for computation-communication overlap.
  ☆219Updated 3 weeks ago
• microsoft / TileFusion
  TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.
  ☆106Updated 7 months ago
• Chtholly-Boss / swizzle
  A practical way of learning Swizzle
  ☆36Updated last year
• xlite-dev / HGEMM
  ⚡️Write HGEMM from scratch using Tensor Cores with WMMA, MMA and CuTe API, Achieve Peak⚡️ Performance.
  ☆148Updated 9 months ago
• antgroup / DeepXTrace
  DeepXTrace is a lightweight tool for precisely diagnosing slow ranks in DeepEP-based environments.
  ☆92Updated 3 weeks ago
• TiledTensor / TiledCUDA
  We invite you to visit and follow our new repository at https://github.com/microsoft/TileFusion. TiledCUDA is a highly efficient kernel …
  ☆192Updated last year
• PipeFusion / PipeFusion
  A Suite for Parallel Inference of Diffusion Transformers (DiTs) on multi-GPU Clusters
  ☆56Updated last year
• tile-ai / tilescale
  Tile-based language built for AI computation across all scales
  ☆120Updated this week
• toyaix / triton-runner
  Multi-Level Triton Runner supporting Python, IR, PTX, and cubin.
  ☆84Updated 2 weeks ago
• microsoft / tokenweave
  Accepted to MLSys 2026
  ☆70Updated 2 weeks ago
• alibaba / easydist
  Automated Parallelization System and Infrastructure for Multiple Ecosystems
  ☆82Updated last year