FlashKDA: high-performance Kimi Delta Attention kernels
☆1,234Jul 30, 2026Updated 3 weeks ago
Alternatives and similar repositories for FlashKDA
Users that are interested in FlashKDA are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- CUDA kernels for linear attention variants, written in CuTe DSL and CUTLASS C++.☆543Updated this week
- high-performance linear attention kernel library built on TileLang☆666Updated this week
- A kernel library written in tilelang☆1,743Apr 23, 2026Updated 4 months ago
- mKernel: fast multi-node, multi-GPU fused kernels☆270Updated this week
- A Quirky Assortment of CuTe Kernels☆1,133Aug 21, 2026Updated last week
- Managed Kubernetes at scale on DigitalOcean • AdDigitalOcean Kubernetes includes the control plane, bandwidth allowance, container registry, automatic updates, and more for free.
- Accelerating MoE with IO and Tile-aware Optimizations☆750Aug 17, 2026Updated last week
- NVIDIA cuTile learn☆168Dec 9, 2025Updated 8 months ago
- ☆37Aug 7, 2025Updated last year
- High-performance LLM operator library built on TileLang.☆173Updated this week
- DeeperGEMM: crazy optimized version☆85May 5, 2025Updated last year
- ☆413Updated this week
- [MLSys 26] 🥇 Solution for Gated Delta Net Track of MLSys 26 Flash infer competition☆36May 22, 2026Updated 3 months ago
- Triton kernels and PyTorch ops for Block Attention Residuals (AttnRes)☆88May 29, 2026Updated 2 months ago
- ☆253Nov 19, 2025Updated 9 months ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- Distributed Compiler and Optimized Parallel Kernels☆1,528Aug 12, 2026Updated 2 weeks ago
- ☆51Aug 17, 2026Updated last week
- TokenSpeed is a speed-of-light LLM inference engine.☆2,026Updated this week
- Learning TileLang with 10 puzzles!☆373Aug 11, 2026Updated 2 weeks ago
- Tile-Based Runtime for Ultra-Low-Latency LLM Inference☆1,739Aug 13, 2026Updated 2 weeks ago
- 🚀 Efficient implementations for emerging model architectures☆5,650Updated this week
- Tilus is a tile-level kernel programming language with explicit control over shared memory and registers.☆490Aug 17, 2026Updated last week
- ☆32Jul 2, 2025Updated last year
- ☆207Updated this week
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- ☆253Updated this week
- Compact and Agent-Native MoE Training System☆343Aug 22, 2026Updated last week
- Skills for writing tilelang and debugging with CUDA toolkits.☆134May 20, 2026Updated 3 months ago
- PTX ISA 9.1 documentation converted to searchable markdown. Includes Claude Code skill for CUDA development.☆227Dec 24, 2025Updated 8 months ago
- FLA but cuTile☆27Apr 17, 2026Updated 4 months ago
- ☆1,592Nov 17, 2025Updated 9 months ago
- FA4-based Relative Attention Kernel developed by TML and Colfax☆18Jul 17, 2026Updated last month
- ☆417Jul 30, 2026Updated 3 weeks ago
- Official repository for Parallax (Parameterized Local Linear Attention)☆68Jul 30, 2026Updated 3 weeks ago
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- a size profiler for cuda binary☆71Jan 15, 2026Updated 7 months ago
- Fast and Memory-Efficient Exact Attention (BF16/FP16/FP8/FP4) for Large Headdim, 1.5x~15x speedup over PyTorch SDPA.☆328Updated this week
- High Performance LLM Inference Operator Library☆1,134Updated this week
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆200Feb 11, 2026Updated 6 months ago
- High-performance GPU kernels written in TIRx.☆93Updated this week
- Distributed MoE in a Single Kernel [NeurIPS '25]☆288May 5, 2026Updated 3 months ago
- An experimental communicating attention kernel based on DeepEP.☆34Jul 29, 2025Updated last year