An experimental communicating attention kernel based on DeepEP.
☆34Jul 29, 2025Updated 10 months ago
Alternatives and similar repositories for AttnLink
Users that are interested in AttnLink are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- ☆66Apr 26, 2025Updated last year
- ☆57Feb 24, 2026Updated 3 months ago
- DeeperGEMM: crazy optimized version☆86May 5, 2025Updated last year
- Benchmark tests supporting the TiledCUDA library.☆19Nov 19, 2024Updated last year
- ☆52May 19, 2025Updated last year
- Virtual machines for every use case on DigitalOcean • AdGet dependable uptime with 99.99% SLA, simple security tools, and predictable monthly pricing with DigitalOcean's virtual machines, called Droplets.
- ☆22May 5, 2025Updated last year
- High performance RMSNorm Implement by using SM Core Storage(Registers and Shared Memory)☆30Jan 22, 2026Updated 4 months ago
- FlashTile is a CUDA Tile IR compiler that is compatible with NVIDIA's tileiras, targeting SM70 through SM121 NVIDIA GPUs.☆61Feb 6, 2026Updated 4 months ago
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆193Feb 11, 2026Updated 4 months ago
- ☆37Aug 7, 2025Updated 10 months ago
- ☆108May 31, 2025Updated last year
- Nex Venus Communication Library☆76Nov 17, 2025Updated 7 months ago
- ☆32Jul 2, 2025Updated 11 months ago
- [WIP] Better (FP8) attention for Hopper☆34Feb 24, 2025Updated last year
- 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.
- DeepXTrace is a lightweight tool for precisely diagnosing slow ranks in DeepEP-based environments.☆99Jan 16, 2026Updated 5 months ago
- ☆20Dec 24, 2024Updated last year
- GEMV implementation with CUTLASS☆21Aug 21, 2025Updated 10 months ago
- Perplexity GPU Kernels☆584Nov 7, 2025Updated 7 months ago
- Debug print operator for cudagraph debugging☆15Aug 2, 2024Updated last year
- Tile-based language built for AI computation across all scales☆168Updated this week
- A high-performance acceleration library dedicated to large-scale model training on AMD GPUs☆64Updated this week
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆532Updated this week
- A NCCL extension library, designed to efficiently offload GPU memory allocated by the NCCL communication library.☆111Dec 17, 2025Updated 6 months ago
- Wordpress hosting with auto-scaling - Free Trial Offer • AdFully Managed hosting for WordPress and WooCommerce businesses that need reliable, auto-scalable performance. Cloudways SafeUpdates now available.
- Accelerate LLM preference tuning via prefix sharing with a single line of code☆53Jul 4, 2025Updated 11 months ago
- Quartet II Official Code☆74May 1, 2026Updated last month
- ☆367Jan 28, 2026Updated 4 months ago
- ⚡️Write HGEMM from scratch using Tensor Cores with WMMA, MMA and CuTe API, Achieve Peak⚡️ Performance.☆156May 10, 2025Updated last year
- Official repository Flash Local Linear Attention☆37May 28, 2026Updated 3 weeks ago
- Low overhead tracing library and trace visualizer for pipelined CUDA kernels☆136Nov 26, 2025Updated 6 months ago
- A Top-Down Profiler for GPU Applications☆22Feb 29, 2024Updated 2 years ago
- A lightweight design for computation-communication overlap.☆237Jan 20, 2026Updated 5 months ago
- Multiple GEMM operators are constructed with cutlass to support LLM inference.☆20Aug 3, 2025Updated 10 months ago
- Wordpress hosting with auto-scaling - Free Trial Offer • AdFully Managed hosting for WordPress and WooCommerce businesses that need reliable, auto-scalable performance. Cloudways SafeUpdates now available.
- Triton adapter for Ascend. Mirror of https://gitcode.com/ascend/triton-ascend☆125May 18, 2026Updated last month
- study of cutlass☆22Nov 10, 2024Updated last year
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆550Updated this week
- 🤖FFPA: Extends FlashAttention-2 via Split-D for large headdims, 1.5x~3×↑🎉 vs SDPA, up to 430T🎉 on H200.☆310Updated this week
- Framework to reduce autotune overhead to zero for well known deployments.☆101Sep 19, 2025Updated 9 months ago
- Quantized Attention on GPU☆44Nov 22, 2024Updated last year
- PyTorch bindings for CUTLASS grouped GEMM.☆153May 29, 2025Updated last year