Summary
AMD has introduced Ross, an AI assistant that connects natural-language requests with AMD embedded development tools, documentation and reusable engineering workflows. The company says it is available now, with further tool and workflow capabilities planned monthly.
AMD has introduced AMD Ross, an agentic AI assistant for embedded-system development. The company says Ross connects natural-language requests to AMD tools, documentation and reusable engineering workflows, covering tasks from hardware design and debugging to software, edge AI and deployment. AMD says Ross is available now and plans to add more embedded tools and workflows monthly.
How Ross connects AI to engineering tools
Ross is designed to do more than answer questions or suggest code. According to AMD, developers can use it to search documentation, check tool status, run commands, generate code, troubleshoot errors and follow established workflows. In practical terms, this makes Ross a natural-language interface to supported development environments: an agent can act on a task through connected tools, rather than returning text alone.
AMD says the system brings together four elements. Model Context Protocol (MCP) servers connect AI agents to AMD embedded tools so they can query information and run commands. An AMD Knowledge Base provides access to guides, white papers, application notes and recorded answers, either through the cloud or in a locally available offline environment. Expert-authored “agent skills”—shareable Markdown files—describe repeatable procedures, while ready-to-run design examples show how those procedures can be applied.
The company lists support across hardware design and debugging, software and algorithm development, edge-AI implementation and deployment. Example tasks include hardware/software partitioning, high-level synthesis, silicon design optimization, power estimation, schematic review and board-layout work. AMD also names tools including Vivado Design Suite, ChipScope, Power Design Manager, Vitis, Vitis HLS, Vitis AI, ROCm AI and Ryzen AI software.
Where the approach could help
Embedded projects can span programmable logic, processors, software and system-level design. AMD’s approach is to make its tools and internal engineering practices accessible through reusable workflows, with the aim of reducing repeated manual interactions and helping teams apply established methods more consistently.
AMD describes example workflows for diagnosing tool errors, investigating timing violations, optimizing Vitis HLS designs and configuring hardware-debug captures. For instance, a timing-analysis workflow can identify violations, explain likely root causes and suggest implementation changes; an HLS workflow can recommend pipelining, loop optimization or pragma changes. These are company-described capabilities and intended uses, rather than quantified performance results in the announcement.
Ross is described as client-agnostic: teams can use their preferred large language models, integrated development environments and command-line environments while connecting to AMD embedded tools. That flexibility may let developers fit the assistant into existing workflows, while the AMD-specific documentation and skills provide context for the company’s own development ecosystem.
Availability and user responsibilities
AMD says Ross is available now through AMD.com/Ross. Additional AMD Embedded tools and workflow capabilities are planned on a monthly cadence.
The company says Ross operates within configured user permissions and supported tool environments. It also places responsibility on users to review generated outputs, recommendations and proposed actions before relying on or implementing them. AMD advises against submitting confidential, personal or regulated data unless its use is authorized for the particular model, tool and deployment configuration.