PKU-SEC-Lab / HybriMoELinks
[DAC'25] Official implement of "HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference"
☆101Updated last month
Alternatives and similar repositories for HybriMoE
Users that are interested in HybriMoE are comparing it to the libraries listed below
Sorting:
- [HPCA 2026] A GPU-optimized system for efficient long-context LLMs decoding with low-bit KV cache.☆80Updated last month
- SpInfer: Leveraging Low-Level Sparsity for Efficient Large Language Model Inference on GPUs☆61Updated 10 months ago
- DLSlime: Flexible & Efficient Heterogeneous Transfer Toolkit☆92Updated 2 weeks ago
- Repo for SpecEE: Accelerating Large Language Model Inference with Speculative Early Exiting (ISCA25)☆70Updated 9 months ago
- Implement Flash Attention using Cute.☆100Updated last year
- Accelerating Large-Scale Reasoning Model Inference with Sparse Self-Speculative Decoding☆87Updated 2 months ago
- A lightweight design for computation-communication overlap.☆219Updated 3 weeks ago
- ☆65Updated 9 months ago
- ☆85Updated 3 months ago
- High performance Transformer implementation in C++.☆151Updated last year
- ☆85Updated last year
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆161Updated 4 months ago
- Keyformer proposes KV Cache reduction through key tokens identification and without the need for fine-tuning☆58Updated last year
- Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity☆233Updated 2 years ago
- ☆130Updated 5 months ago
- A collection of specialized agent skills for AI infrastructure development, enabling Claude Code to write, optimize, and debug high-perfo…☆54Updated last week
- NVIDIA cuTile learn☆158Updated 2 months ago
- DeeperGEMM: crazy optimized version☆73Updated 9 months ago
- FlexFlow Serve: Low-Latency, High-Performance LLM Serving☆71Updated 4 months ago
- ArkVale: Efficient Generative LLM Inference with Recallable Key-Value Eviction (NIPS'24)☆53Updated last year
- Nex Venus Communication Library☆72Updated 2 months ago
- ☆164Updated last year
- Triton adapter for Ascend. Mirror of https://gitee.com/ascend/triton-ascend☆107Updated this week
- A prefill & decode disaggregated LLM serving framework with shared GPU memory and fine-grained compute isolation.☆123Updated last month
- Building the Virtuous Cycle for AI-driven LLM Systems☆164Updated this week
- Tile-based language built for AI computation across all scales☆120Updated this week
- PyTorch library for cost-effective, fast and easy serving of MoE models.☆279Updated last week
- Explore Inter-layer Expert Affinity in MoE Model Inference☆16Updated last year
- ☆58Updated last year
- 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