A new AI-native development experience empowers developers to install, deploy, troubleshoot, and optimize AI workloads on AMD platforms using natural language and leading AI coding assistants.
Key Highlights:
1) At the Advancing AI 2026 event, AMD launched ROCm.ai, an AI-native software experience designed for AI development on AMD platforms.
2) ROCm.ai consolidates the ROCm CLI, AMD Skills, and AI-driven software optimization capabilities into a unified developer experience.
3) The platform supports developers in using natural language and AI coding assistants to install, deploy, troubleshoot, and optimize AI workloads.
4) Through AI-driven optimization of compute kernels, memory management, parallelization, and scheduling mechanisms, ROCm.ai further refines the ROCm software stack and boosts development efficiency.
5) On the same hardware platform, compared to ROCm 7, the latest version of ROCm with ROCm.ai achieved an average 3.3x improvement in inference performance and a 2.4x improvement in training performance in tests.
Content Summary:
AMD unveiled ROCm.ai at the Advancing AI 2026 conference. This AI-native development experience combines the ROCm CLI, AMD Skills, and AI-driven software optimization capabilities to help developers build, deploy, and optimize AI workloads on AMD platforms. The platform integrates AMD's expertise into leading AI coding assistants like Claude, Codex, and Cursor. On the same hardware platform, compared to ROCm 7, the latest version of ROCm with ROCm.ai achieved an average 3.3x inference performance gain and a 2.4x training performance gain in tests.
What is ROCm.ai?
ROCm.ai is AMD's AI-native development experience for AI development, designed to help developers build, deploy, and optimize AI workloads on AMD platforms.
What components make up ROCm.ai?
ROCm.ai integrates the ROCm CLI, AMD Skills, and AI-driven software optimization capabilities, including Hyperloom, into a unified software experience.
Why did AMD launch ROCm.ai?
AMD launched ROCm.ai to help developers quickly translate development ideas into production applications using natural language and AI coding assistants.
How does ROCm.ai improve performance?
ROCm.ai enhances AI inference and training performance by applying AI-driven optimization techniques to compute kernels, memory management, parallelization, and scheduling mechanisms.
What are the performance improvements announced by AMD?
AMD stated that, compared to AMD ROCm 7 on the same hardware platform, the latest version of ROCm with ROCm.ai achieved an average 3.3x improvement in inference performance and a 2.4x improvement in training performance in tests.
What are the noteworthy new developments?
AMD officially announced the release of ROCm.ai, an AI-native software development experience for AMD platforms. ROCm.ai consolidates AI-assisted development, intelligent deployment, and AI-driven software optimization into a unified software experience, helping developers more quickly move from development concepts to production deployment. ROCm.ai was officially unveiled at the Advancing AI 2026 event. Its components include the ROCm CLI, a new command-line tool for installing, verifying, deploying, and managing AI workloads; AMD Skills, which integrates official AMD technical knowledge into leading AI coding assistants like Claude, Cursor, and Codex; and Hyperloom, a new open-source Agentic AI system designed to automate the time-consuming engineering task of end-to-end inference workload optimization. Through this unified software experience, ROCm.ai helps developers install, deploy, troubleshoot, and optimize AI workloads on AMD platforms.
Why is this significant?
AI development is increasingly moving toward an Agent-driven model, with more developers using natural language and AI coding assistants to build, deploy, and optimize software. ROCm.ai integrates AMD's technical capabilities directly into these development workflows, providing developers with expert guidance specific to AMD platforms, rather than generic advice, to help them build and optimize AI workloads more efficiently. The introduction of AMD Skills further lowers the barrier to entry for developers using ROCm and reduces the time spent consulting documentation during environment configuration. ROCm.ai also uses AI to continuously optimize the AMD ROCm software stack. Through AI-driven optimization of compute kernels, memory management, and scheduling mechanisms, the latest version of ROCm, with ROCm.ai, achieved an average 3.3x improvement in inference performance and a 2.4x improvement in training performance on the same hardware platform compared to ROCm 7.
What role did AMD play in this?
"AMD is committed to helping developers work with unprecedented efficiency. ROCm.ai brings an Agentic AI development experience to AMD platforms. These AI Agents can not only answer questions but also analyze performance, debug applications, and drive workloads to fully unleash performance on AMD hardware. This marks a new phase of development for the AMD AI software stack, further shortening the distance between development concepts and actual operation, enabling developers to complete development and iteration faster. Whether setting up a development environment, deploying models, or further optimizing performance, ROCm.ai can accomplish these key tasks within the AI coding assistants that developers use daily." — Anush Elangovan, Corporate Vice President, AI Software, AMD
What is ROCm.ai?
ROCm.ai is AMD's AI-native development experience for AI development, designed to help developers develop, deploy, and optimize AI applications on AMD platforms more quickly. It integrates the ROCm CLI, AMD Skills, and AI-driven software optimization capabilities into a unified software experience, simplifying the entire AI development process from installation and deployment to debugging and performance optimization.
The ROCm CLI provides a unified command-line experience for installing, verifying, deploying, updating, and troubleshooting AI workloads on AMD platforms. It features hardware-aware workflows, automated environment configuration, and support for air-gapped deployments, helping to simplify the software installation process and ensure consistent deployment across environments.
AMD Skills integrates official AMD technical knowledge into leading AI coding assistants like Claude, Cursor, and Codex. Instead of relying on general advice, developers can receive guidance on installation, migration, debugging, and performance optimization specific to the AMD AI software stack.
Hyperloom automates the end-to-end optimization of AI inference workloads, covering the complete process from performance analysis and bottleneck diagnosis to compute kernel optimization and result validation. This helps developers complete optimization tasks that would typically require weeks of specialized engineering effort within hours.
Through AI-driven software optimization capabilities, including Hyperloom, ROCm.ai continuously optimizes the ROCm software stack in areas such as parallelization and task scheduling, memory management, and compute kernel optimization. These software optimizations further enhance inference and training performance on existing AMD hardware and continuously improve development efficiency and workload performance.
ROCm.ai will be available for use starting in August 2026.
Comments