A developer has presented a project for designing a RISC-V processor, a System on Chip (SoC), and a printed circuit board (PCB) using the capabilities of the AI assistant Claude.

What Happened

The project demonstrates the use of Claude for analyzing RISC-V specifications, writing SystemVerilog code, and designing boards via the SKiDL tool. The author also published a repository containing an agentic environment and build tools under the MIT license.

Context

LLMs are beginning to move beyond purely text and software tasks, integrating with Electronic Design Automation (EDA) tools. In this case, Claude acts as an orchestrator between high-level specifications and specialized Hardware Description Languages (HDL).

Why It Matters for the Industry

This technology demonstrates the potential for automating EDA processes, which could significantly lower the barrier to entry for hardware startups and simplify the creation of specialized hardware (ASIC/FPGA), turning complex engineering tasks into AI-agent management processes.

Why It Matters for Users

Hardware enthusiasts and AI developers gain access to ready-made agent templates and automation tools, allowing them to quickly create working prototypes of their own CPUs and SoCs.

What Is Not Yet Known / Limitations

Current solutions serve more as powerful prototyping tools rather than full replacements for production-grade industrial solutions.

Sources

Author

Look at AI, Editorial Team