hipObject: RDMA-Accelerated S3 Object Storage for GPUs

Warning

hipObject is an experimental library. It is not intended for production use.

Introduction

hipObject enables direct data transfers between AMD GPU VRAM and S3-compatible object storage using RDMA over RoCEv2. It interoperates with NVIDIA cuObject-equipped storage servers via the x-amz-rdma-token S3 header protocol, providing a vendor-neutral client for GPU-direct object storage.

Architecture

hipObject separates control and data planes:

• Control plane: Standard S3 REST requests augmented with x-amz-rdma-token / x-amz-rdma-reply headers.

• Data plane: RDMA READ/WRITE over RoCEv2 RC transport directly between GPU VRAM and storage server buffers.

Supported S3 Operations

Operation	Description
GET	Fetch object to GPU VRAM
PUT	Store GPU VRAM to object
UPLOAD	Chunked upload (multipart)
RANGE	Byte-range fetch from object

User Guide

• Building hipObject
  • Prerequisites
  • Build Steps
  • CMake Options
  • Packaging
  • MinIO C++ RDMA Bridge
  • cuObject Interoperability
  • Building Documentation
  • Environment Variables
• hipObject cuObject Interoperability
  • Compatibility matrix
  • Testing paths
  • Hardware topology
  • CMake options

API Reference

• API Reference
  • Core Functionality
  • Thread Safety
  • Errors and Error Handling
  • Buffer Registration
  • Data Transfer (GET / PUT)

License

MIT License