uccl-project / ucclLinks
UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)
☆1,066Updated last week
Alternatives and similar repositories for uccl
Users that are interested in uccl are comparing it to the libraries listed below
Sorting:
- Extending eBPF Programmability and Observability to GPUs (merged into https://github.com/eunomia-bpf/bpftime)☆267Updated last week
- ☆736Updated 3 weeks ago
- Unified KV Cache Compression Methods for Auto-Regressive Models☆1,277Updated 10 months ago
- An acceleration library that supports arbitrary bit-width combinatorial quantization operations☆238Updated last year
- A highly optimized LLM inference acceleration engine for Llama and its variants.☆904Updated 4 months ago
- CXL remote offloading data movement aware compiler☆30Updated last month
- [Neurips 2025] R-KV: Redundancy-aware KV Cache Compression for Reasoning Models☆1,148Updated last month
- ☆918Updated this week
- TVM Documentation in Chinese Simplified / TVM 中文文档☆2,653Updated 2 weeks ago
- [ICLR 2025🔥] SVD-LLM & [NAACL 2025🔥] SVD-LLM V2☆262Updated 2 months ago
- Heterogeneous Containerization of Large Language Model Apps☆107Updated 3 months ago
- NVIDIA Inference Xfer Library (NIXL)☆721Updated last week
- Expert Kit is an efficient foundation of Expert Parallelism (EP) for MoE model Inference on heterogenous hardware☆59Updated 3 weeks ago
- Efficient and easy multi-instance LLM serving☆510Updated 2 months ago
- [ICML 2025 Spotlight] ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference☆272Updated 6 months ago
- MIXQ: Taming Dynamic Outliers in Mixed-Precision Quantization by Online Prediction☆94Updated last year
- FlagPerf is an open-source software platform for benchmarking AI chips.☆353Updated 2 weeks ago
- Some Hardware Architectures for GEMM☆283Updated 6 months ago
- PTX on XPUs☆109Updated last week
- ☆320Updated last week
- [COLM 2024] TriForce: Lossless Acceleration of Long Sequence Generation with Hierarchical Speculative Decoding☆271Updated last year
- Perplexity GPU Kernels☆529Updated 2 weeks ago
- Dynamic Memory Management for Serving LLMs without PagedAttention☆436Updated 5 months ago
- torchcomms: a modern PyTorch communications API☆291Updated this week
- The official implementation of MARS: Unleashing the Power of Variance Reduction for Training Large Models☆713Updated 3 weeks ago
- KV cache store for distributed LLM inference☆363Updated last week
- Intelligent Router for Mixture-of-Models☆2,294Updated this week
- Disaggregated serving system for Large Language Models (LLMs).☆734Updated 7 months ago
- A low-latency & high-throughput serving engine for LLMs☆447Updated last month
- Fast OS-level support for GPU checkpoint and restore☆257Updated last month