Dao-AILab / sonic-moeView external linksLinks
Accelerating MoE with IO and Tile-aware Optimizations
☆583Feb 6, 2026Updated last week
Alternatives and similar repositories for sonic-moe
Users that are interested in sonic-moe are comparing it to the libraries listed below
Sorting:
- Distributed Compiler based on Triton for Parallel Systems☆1,350Updated this week
- Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels☆5,094Feb 7, 2026Updated last week
- ☆118May 19, 2025Updated 8 months ago
- Accelerating Large-Scale Reasoning Model Inference with Sparse Self-Speculative Decoding☆88Dec 2, 2025Updated 2 months ago
- FlashInfer: Kernel Library for LLM Serving☆4,935Updated this week
- 🚀 Efficient implementations of state-of-the-art linear attention models☆4,379Updated this week
- LM engine is a library for pretraining/finetuning LLMs☆115Updated this week
- 🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"☆964Feb 5, 2026Updated last week
- Tile primitives for speedy kernels☆3,139Updated this week
- Fast and memory-efficient exact kmeans☆138Updated this week
- ☆38Aug 7, 2025Updated 6 months ago
- Perplexity GPU Kernels☆560Nov 7, 2025Updated 3 months ago
- Nex Venus Communication Library☆72Nov 17, 2025Updated 2 months ago
- PyTorch bindings for CUTLASS grouped GEMM.☆184Dec 16, 2025Updated last month
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,247Aug 28, 2025Updated 5 months ago
- A sparse attention kernel supporting mix sparse patterns☆455Jan 18, 2026Updated 3 weeks ago
- A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on H…☆3,152Feb 7, 2026Updated last week
- FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.☆1,011Sep 4, 2024Updated last year
- DeeperGEMM: crazy optimized version☆73May 5, 2025Updated 9 months ago
- A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.☆744Updated this week
- 🤖FFPA: Extend FlashAttention-2 with Split-D, ~O(1) SRAM complexity for large headdim, 1.8x~3x↑🎉 vs SDPA EA.☆251Updated this week
- ☆129Jun 6, 2025Updated 8 months ago
- Efficient Long-context Language Model Training by Core Attention Disaggregation☆89Jan 29, 2026Updated 2 weeks ago
- incubator repo for CUDA-TileIR backend☆102Updated this week
- Tile-Based Runtime for Ultra-Low-Latency LLM Inference☆567Jan 26, 2026Updated 2 weeks ago
- A Quirky Assortment of CuTe Kernels☆798Updated this week
- ☆19Aug 20, 2025Updated 5 months ago
- Mirage Persistent Kernel: Compiling LLMs into a MegaKernel☆2,120Jan 29, 2026Updated 2 weeks ago
- ☆86Updated this week
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆462Dec 31, 2025Updated last month
- Helpful kernel tutorials and examples for tile-based GPU programming☆641Updated this week
- [ICLR 2025] DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads☆524Feb 10, 2025Updated last year
- [ICML 2024] Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference☆372Jul 10, 2025Updated 7 months ago
- ☆221Nov 19, 2025Updated 2 months ago
- [ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-t…☆3,159Jan 17, 2026Updated 3 weeks ago
- a size profiler for cuda binary☆72Jan 15, 2026Updated last month
- DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling☆6,162Feb 3, 2026Updated last week
- Analyze computation-communication overlap in V3/R1.☆1,141Mar 21, 2025Updated 10 months ago
- [MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Se…☆812Mar 6, 2025Updated 11 months ago