nvidia-nixl

Official

High-performance cross-node KV transfers.

Authorair-gapped
Version1.0.0
Installs0

System Documentation

What problem does it solve?

NVIDIA Inference Xfer Library (NIXL) provides a transport layer for high-performance cross-node memory transfers and metadata coordination in distributed inference stacks.

Core Features & Use Cases

  • Pluggable backends across UCX, libfabric, GDS, POSIX, HF3FS, and more to cover various network/storage fabrics.
  • Metadata exchange via side-channel TCP or ETCD for elastic clusters, enabling Dynamo/vLLM/SGLang deployments to scale.
  • Transfer lifecycle primitives: register memory, descriptor lists, initialize_xfer, post, check_xfer_state, and telemetry for observability.

Quick Start

Launch two processes acting as target and initiator and perform a minimal two-peer tensor transfer using the Python API.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: nvidia-nixl
Download link: https://github.com/air-gapped/skills/archive/main.zip#nvidia-nixl

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 510,000+ vetted skills library on demand.