ai-dynamo / nixlLinks
NVIDIA Inference Xfer Library (NIXL)
☆788Updated last week
Alternatives and similar repositories for nixl
Users that are interested in nixl are comparing it to the libraries listed below
Sorting:
- Efficient and easy multi-instance LLM serving☆520Updated 4 months ago
- Perplexity GPU Kernels☆548Updated 2 months ago
- KV cache store for distributed LLM inference☆384Updated last month
- Virtualized Elastic KV Cache for Dynamic GPU Sharing and Beyond☆735Updated last month
- Disaggregated serving system for Large Language Models (LLMs).☆761Updated 9 months ago
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆449Updated this week
- A low-latency & high-throughput serving engine for LLMs☆464Updated this week
- Offline optimization of your disaggregated Dynamo graph☆137Updated this week
- NVIDIA Resiliency Extension is a python package for framework developers and users to implement fault-tolerant features. It improves the …☆247Updated this week
- Dynamic Memory Management for Serving LLMs without PagedAttention☆454Updated 7 months ago
- Materials for learning SGLang☆714Updated last week
- torchcomms: a modern PyTorch communications API☆319Updated this week
- GLake: optimizing GPU memory management and IO transmission.☆494Updated 9 months ago
- Byted PyTorch Distributed for Hyperscale Training of LLMs and RLs☆917Updated last month
- Distributed Compiler based on Triton for Parallel Systems☆1,307Updated 2 weeks ago
- A tool for bandwidth measurements on NVIDIA GPUs.☆602Updated 8 months ago
- CUDA checkpoint and restore utility☆401Updated 3 months ago
- Genai-bench is a powerful benchmark tool designed for comprehensive token-level performance evaluation of large language model (LLM) serv…☆251Updated this week
- Open Model Engine (OME) — Kubernetes operator for LLM serving, GPU scheduling, and model lifecycle management. Works with SGLang, vLLM, T…☆355Updated this week
- Zero Bubble Pipeline Parallelism☆445Updated 8 months ago
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,218Updated 4 months ago
- A throughput-oriented high-performance serving framework for LLMs☆936Updated 2 months ago
- ☆338Updated last week
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆433Updated last week
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆615Updated this week
- NVIDIA NCCL Tests for Distributed Training☆132Updated this week
- Perplexity open source garden for inference technology☆324Updated 2 weeks ago
- NCCL Fast Socket is a transport layer plugin to improve NCCL collective communication performance on Google Cloud.☆122Updated 2 years ago
- Serverless LLM Serving for Everyone.☆628Updated last week
- A large-scale simulation framework for LLM inference☆513Updated 5 months ago