AI agents are all the rage, and the technology is being integrated into embedded development environments to boost productivity. AI can be used for code generation and configuration of platforms, and even for the design of hardware in FPGAs, but there are issues with the quality and the safety of code that is produced. “Natural language as the new abstraction layer for embedded in both the hardware and software domains,” said Kirk Saban, vice president for Product, Software & Solutions at AMD. “The fundamental game changer has been the advent of large language models (LLMs) with the ability to describe the problem in natural language.” “One of the key things we see in embedded is that we deal with all different types of domains, form factors, use cases, and that requires a lot of different disciplines and architectural trade-offs,” Saban tells eeNews Europe . “There’s often many manual handoffs and there are lots of different tools out there and as the chips have become more complicated so have the tools.” The result for AMD is a harness called Ross that acts as an interface and allows AMD tools to communicate across a range of tools. Ross, named in the same way as Anthropic’s Claude AI, is named after Ross Freeman, inventor of the Field-Programmable Gate Array (FPGA) and former director of engineering at Zilog and a co-founder of Xilinx, alongside Bernie Vonderschmidt and James Burnett. “ AMD Ross spans the entire paradigm of our development stack, running alongside and plugging into all the different tools,” said Saban. This works across the range of AMD tools, from the Vivado FPGA design tools and Vitis unified IDE development environment to the Ryzen embedded processor AI tools and the ChipScope integrated logic analyzer. The Ross agentic AI harness from AMD “We are not in the business of creating LLMs – Ross will plug into any of the public model providers, but we also support open weight models. It’s a harness across our entire development stack. We are giving a harness for you to adapt to your design environment,” said Saban. “It is the same with IDE tools, we work with all the major players in the industry.” AdaCore is also using its GNAT Pro development environment for AI-generated embedded code for safety-critical systems using the Ada and SPARK formal languages and the GNAT Pro development environment. “We know that Spark requires an additional level of effort. You can write it without the additional formal methods, but it is harder,” said Mark Hermeling, head of technical marketing at AdaCore. “But if you have Ada code and you want to lift that to SPARK Silver [to guarantee the absence of runtime errors], AI can do that in an LLM, every time. This is a tremendous offload for the developer.” He took a standard STMicroelectronics demo with some C code for data gathering, translated it to Ada and lifted it to SPARK Silver, all on a laptop on an aircraft travelling from Los Angeles to Toronto. It gave six areas to inspect incoming data and can also set select functional properties, typically for safety or security in the next level, SPARK Gold, SPARK Platinum cover the full formally proven design for safety critical code such as automotive, aerospace or encryption. “So whatever AI does it can never violate those requirements, so it’s really easy for developers to elaborate the code while staying completely safe,” he said. The AI support is provided by a repository of skills on Github. “This includes what we have learned over the last six to nine months,” he said. “We started with Model Context Protocol (MCP) but this is not necessary for command line tools, the skills do that. Specifically, the Connect skill codifies SPARK best practices. Grab the skills from the repo and use the Prove skills. This spawns a set of agents in a particular pattern to lift the Ada code to SPARK Silver,” he said. “What GNAT Foundry provides is a workflow for agents to follow the V-model for requirements. There’s an agentic framework with di