> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/nixl/user-guide/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nixl/_mcp/server.

# User Guide

## Docs

- [Backend Selection](https://docs.nvidia.com/nixl/user-guide/backend-selection.md): How to choose the right NIXL transfer backend for your memory types and use case.
- [UCX](https://docs.nvidia.com/nixl/user-guide/backend-selection/ucx.md): UCX transfer backend for high-performance VRAM and DRAM transfers via RDMA and TCP.
- [Libfabric](https://docs.nvidia.com/nixl/user-guide/backend-selection/libfabric.md): Libfabric (OFI) network backend for VRAM-to-VRAM, VRAM-to-DRAM, and DRAM-to-DRAM transfers as an alternative to UCX.
- [Mooncake](https://docs.nvidia.com/nixl/user-guide/backend-selection/mooncake.md): Mooncake Transfer Engine backend for VRAM and DRAM transfers in Mooncake-enabled deployments.
- [UCCL-P2P](https://docs.nvidia.com/nixl/user-guide/backend-selection/uccl.md): UCCL-P2P network backend for VRAM and DRAM transfers.
- [DOCA GPUNetIO](https://docs.nvidia.com/nixl/user-guide/backend-selection/gpunetio.md): DOCA GPUNetIO backend for high-performance GPU-to-GPU transfers using GPUDirect Async.
- [GDS](https://docs.nvidia.com/nixl/user-guide/backend-selection/gds.md): GPUDirect Storage backend for direct GPU-to-file transfers without CPU bounce buffers.
- [GDS_MT](https://docs.nvidia.com/nixl/user-guide/backend-selection/gds-mt.md): Multi-threaded GPUDirect Storage backend for higher throughput on parallel file operations.
- [POSIX](https://docs.nvidia.com/nixl/user-guide/backend-selection/posix.md): Standard POSIX I/O backend for DRAM-to-file transfers.
- [HF3FS](https://docs.nvidia.com/nixl/user-guide/backend-selection/hf3fs.md): HuggingFace 3FS backend for file operations on 3FS fuse-based filesystems.
- [OBJ](https://docs.nvidia.com/nixl/user-guide/backend-selection/obj.md): Object storage backend supporting S3-compatible stores with multiple executor variants.
- [Azure Blob](https://docs.nvidia.com/nixl/user-guide/backend-selection/azure-blob.md): Azure Blob Storage backend for DRAM-to-object-storage transfers via Azure REST API.
- [INFINIA](https://docs.nvidia.com/nixl/user-guide/backend-selection/infinia.md): DDN INFINIA backend for asynchronous transfers between DRAM or VRAM and object storage.
- [GUSLI](https://docs.nvidia.com/nixl/user-guide/backend-selection/gusli.md): Block-level I/O backend for transfers between block storage devices and DRAM.
- [Metadata Exchange with etcd](https://docs.nvidia.com/nixl/user-guide/metadata-exchange-with-etcd.md): Configure etcd-based distributed metadata exchange for NIXL agents -- key prefix scheme, watcher lifecycle, and Prometheus integration.
- [Telemetry Guide](https://docs.nvidia.com/nixl/user-guide/telemetry-guide.md): Understanding NIXL's telemetry system -- event types, metrics, configuration, and monitoring with Prometheus.
- [Logging Guide](https://docs.nvidia.com/nixl/user-guide/logging-guide.md): Configuring NIXL's logging -- verbosity, per-process log files, filename templates, size limits, and failure behavior.
- [NIXLBench](https://docs.nvidia.com/nixl/user-guide/benchmarking-nixl/nixl-bench.md): A benchmarking tool for measuring NIXL data transfer performance across network and storage backends.
- [Building NIXLBench](https://docs.nvidia.com/nixl/user-guide/benchmarking-nixl/nixl-bench/building-nixl-bench.md): Build NIXLBench using Docker or natively from source with Meson.
- [NIXLBench Usage and Troubleshooting](https://docs.nvidia.com/nixl/user-guide/benchmarking-nixl/nixl-bench/using-nixl-bench.md): How to run NIXLBench benchmarks and troubleshoot common issues.
- [KVBench](https://docs.nvidia.com/nixl/user-guide/benchmarking-nixl/kv-bench.md): A KV cache benchmarking utility that generates and runs NIXLBench commands to profile LLM inference transfer performance.
- [Building KVBench](https://docs.nvidia.com/nixl/user-guide/benchmarking-nixl/kv-bench/building-kv-bench.md): Install KVBench using the NIXLBench Docker container or a Python virtual environment.
- [KVBench Commands and Examples](https://docs.nvidia.com/nixl/user-guide/benchmarking-nixl/kv-bench/using-kv-bench.md): KVBench command reference, model configuration guide, and LLM example configurations.