CUDA kernels for linear attention variants, written in CuTe DSL and CUTLASS C++.
☆533Jul 18, 2026Updated this week
Alternatives and similar repositories for cuLA
Users that are interested in cuLA are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- FlashKDA: high-performance Kimi Delta Attention kernels☆455May 26, 2026Updated last month
- high-performance linear attention kernel library built on TileLang☆597Updated this week
- A Quirky Assortment of CuTe Kernels☆1,060Updated this week
- A kernel library written in tilelang☆1,642Apr 23, 2026Updated 2 months ago
- Accelerating MoE with IO and Tile-aware Optimizations☆731Jul 4, 2026Updated 2 weeks ago
- Proton VPN Special Offer - Get 70% off • AdSpecial partner offer. Trusted by over 100 million users worldwide. Tested, Approved and Recommended by Experts.
- Skills for writing tilelang and debugging with CUDA toolkits.☆131May 20, 2026Updated last month
- ☆250Nov 19, 2025Updated 8 months ago
- 🚀 Efficient implementations for emerging model architectures☆5,367Updated this week
- mKernel: fast multi-node, multi-GPU fused kernels☆251Jun 21, 2026Updated 3 weeks ago
- ☆309Jun 9, 2026Updated last month
- High Performance LLM Inference Operator Library☆1,036Jul 2, 2026Updated 2 weeks ago
- Distributed MoE in a Single Kernel [NeurIPS '25]☆272May 5, 2026Updated 2 months ago
- ☆32Jul 2, 2025Updated last year
- TokenSpeed is a speed-of-light LLM inference engine.☆1,629Updated this week
- End-to-end encrypted cloud storage - Proton Drive • AdSpecial offer: 40% Off Yearly / 80% Off First Month. Protect your most important files, photos, and documents from prying eyes.
- PTX ISA 9.1 documentation converted to searchable markdown. Includes Claude Code skill for CUDA development.☆215Dec 24, 2025Updated 6 months ago
- Learning TileLang with 10 puzzles!☆338May 28, 2026Updated last month
- ☆754Jun 2, 2026Updated last month
- A benchmark of real-world DL kernel problems☆257Updated this week
- CODA: Rewriting Transformer Blocks as GEMM-Epilogue Programs☆229Updated this week
- Compact and Agent-Native MoE Training System☆288Updated this week
- Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels☆6,667Updated this week
- A high-performance kernel library for LLM training☆86Apr 28, 2026Updated 2 months ago
- High-performance LLM operator library built on TileLang.☆161Updated this week
- Managed Database hosting by DigitalOcean • AdPostgreSQL, MySQL, MongoDB, Kafka, Valkey, and OpenSearch available. Automatically scale up storage and focus on building your apps.
- FlashInfer: Kernel Library for LLM Serving☆5,983Updated this week
- High performance RMSNorm Implement by using SM Core Storage(Registers and Shared Memory)☆30Jan 22, 2026Updated 5 months ago
- NVIDIA cuTile learn☆169Dec 9, 2025Updated 7 months ago
- Tile-Based Runtime for Ultra-Low-Latency LLM Inference☆1,573Updated this week
- Helpful kernel tutorials, examples and SKILLs for tile-based GPU programming☆773Updated this week
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆195Feb 11, 2026Updated 5 months ago
- ☆48Updated this week
- An Optimizer for Nvidia Compilers.☆107Jul 3, 2026Updated 2 weeks ago
- Some funny cute/cuteDSL code snippets☆33Mar 2, 2026Updated 4 months ago
- Bare Metal GPUs on DigitalOcean Gradient AI • AdPurpose-built for serious AI teams training foundational models, running large-scale inference, and pushing the boundaries of what's possible.
- Open ABI and FFI for Machine Learning Systems☆433Updated this week
- FLA but cuTile☆27Apr 17, 2026Updated 3 months ago
- FlashInfer Bench @ MLSys 2026: Building AI agents to write high performance GPU kernels☆175Apr 26, 2026Updated 2 months ago
- Distributed Compiler based on Triton for Parallel Systems☆1,493Jul 11, 2026Updated last week
- ☆229Updated this week
- A Distributed Attention Towards Linear Scalability for Ultra-Long Context, Heterogeneous Data Training☆882Updated this week
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆560Updated this week