Hyperloom documentation

2026-08-27

1 min read time

Applies to Linux

ROCm Hyperloom is an autonomous agentic system designed to optimize end-to-end inference workloads on AMD GPUs. It uses advanced AI agents and profiling tools to analyze a workload, identify performance bottlenecks, implement targeted optimizations, and validate the performance and correctness of the optimizations – all without requiring manual intervention.

The Hyperloom source code is hosted on GitHub at AMD-AGI/Hyperloom.

Install

• Install on Docker or bare metal

• Install on a Slurm cluster

Components

• Components

• Hyperloom optimization loop

How to

• Run a Hyperloom optimization

• Optimize your own workload

• Quantization with AMD Quark

Reference

• API reference

• Environment variables

• Authentication and credentials

• Kernel optimization execution path

• Operations and self-hosting

• Upgrade guide

• Session output schema

• Knowledge base integration

• Operator scripts

• Multi-node inference optimization

• Troubleshooting

To contribute to the documentation, see Contributing to Hyperloom.

Hyperloom is released under the MIT license. For details, see the License page.