ROCm Compute Profiler documentation

This documentation provides a comprehensive overview of the ROCm Compute Profiler tool. In addition to a full deployment guide with installation instructions, this documentation also explains the ideas motivating the design behind the tool and its components.

If you’re new to ROCm Compute Profiler, familiarize yourself with the tool by reviewing the chapters that follow and gradually learn its more advanced features. To get started, see What is ROCm Compute Profiler?.

ROCm Compute Profiler is open source and hosted at ROCm/rocprofiler-compute.

Install

• Installation and deployment

• Grafana server for ROCm Compute Profiler

Use the following topics to learn more about the advantages of ROCm Compute Profiler in your development toolkit, how it aims to model performance, and how to use ROCm Compute Profiler in practice.

How to

• Basic usage

• Profile mode

• Analyze mode

  • CLI analysis

  • Grafana GUI analysis

  • Standalone GUI analysis

Conceptual

• Performance model

  • Compute unit (CU)

  • L2 cache (TCC)

  • Shader engine (SE)

  • Command processor (CP)

  • System Speed-of-Light

• Definitions

  • Normalization units

Tutorials

• Profiling by example

• Learning resources

Reference

• Compatible accelerators

• FAQ

This project is proudly open source. For more details on how to contribute, refer to Contributing to ROCm.

Find ROCm licensing information on the Licensing page.