Accelerating Vitis HLS Development with the AMD Ross Agentic AI Assistant
Course Details
Length:
12 Hours
Number of Labs:
3
Number of Chapter:
4
Current Version:
2026.1
Number of Demos:
0
Overview
This course introduces the AMD Ross™ agentic AI assistant for Vitis™ high-level synthesis (HLS) and demonstrates how AI-assisted workflows can improve productivity throughout the HLS development process. Participants examine common challenges in traditional HLS development, explore the concepts of model context protocol (MCP), grounded knowledge, and agent skills, and learn how the Ross assistant supports architecture generation, optimization, and workflow execution. Through a series of hands-on labs, attendees gain practical experience using these AI-assisted workflows to accelerate development while maintaining engineering review and validation.
The emphasis of this course is on:
Understanding common challenges in traditional HLS development
Exploring the concepts of MCP, grounded knowledge, and agent skills
Using the Ross agentic AI assistant for architecture generation, optimization, and workflow execution
Gaining hands-on experience with Ross-assisted HLS workflows across three labs
Maintaining engineering review and validation within the Ross-assisted flow
License requirement:To complete the hands-on labs, participants must have their own access to a supported AI provider (for example, an Anthropic or Open AI account or API key) for use with the Claude Code CLI and Vitis Theia AI.
Note:AI-assisted workflows are model-dependent. Some models may require multiple iterations and prompt refinements before reaching the desired outcome. Models with larger context windows and stronger reasoning capabilities generally achieve the expected results more consistently and with fewer interactions.
