AMD has introduced AMD Ross, an agentic AI assistant built to speed up the full embedded development flow, from hardware and software design through optimisation, debugging, AI implementation and deployment.
Ross is purpose-built for AMD Embedded development environments. It pairs direct access to AMD tools with a grounded AMD Knowledge Base and a set of expert-authored agent skills, with the stated aim of helping engineers automate repetitive workflows and move from design intent to a working implementation faster.
What AMD Ross actually does
Rather than acting as a general chatbot, Ross is positioned as a development companion that stays connected to AMD Embedded tools while the engineer works. AMD says it supports developers' preferred large language models, integrated development environments and command-line environments, so teams are not forced to abandon the setup they already use.
The agentic label is the part that matters. Where an ordinary assistant answers a question, an agentic one is designed to carry out multi-step tasks on the developer's behalf, chaining actions across design, debugging and deployment. It is the same shift we have been tracking across the industry, from Perplexity running large models locally on AMD Ryzen AI Max PCs to a plain-language explainer on what agentic AI actually is.
Why it matters
Embedded development covers the chips and software inside everything from industrial controllers to medical devices and cars, and it is traditionally slow, highly specialised work. A tool that automates parts of that flow is aimed squarely at shortening the gap between an idea and shipping silicon.
Agentic tools are also moving from pilots into real deployments closer to home: Malaysia's own government portal is already testing agentic AI for public services. AMD has not announced Malaysia-specific availability or pricing for Ross, and full technical details sit on the AMD newsroom.
