Efabless Corporation, the creator platform for chips, today announced the winners of its AI Generated Open-Source Silicon Design Challenge. The challenge invited participants to use generative AI to design and tapeout an open-source silicon chip and to do so in three weeks. The wide range of participants and the variety of innovative designs showed how generative AI can accelerate and democratize innovation in chips.
The first-place winner of the contest, by the slimmest of margins, is QTCore-C1 by Hammond Pearce at New York University. The design is a co-processor that can be used for many applications, such as predictable-time I/O state machines for PIO functions as seen on some microcontrollers developed using the Chip-Chat methodology that the NYU team has published.
The second-place winner of the contest is Cyberrio by Xinze Wang, Guohua Yin, and Yifei Zhu at Tsinghua-Berkeley Shenzhen Institute. This design is a RISC-V CPU, implemented with Verilog code produced via a series of prompts given to ChatGPT-4.
The third-place winner of the contest is Asma Mohsin at Rapid Silicon. The design is a Model Predictive Control (MPC) that is used to predict future behavior and optimize control actions for a regulator control circuit provided in MATLAB code to ChatGPT-4 and then implemented with prompts in Verilog.
The winners and all the participants successfully demonstrated how tools such as ChatGPT, Bard and others can revolutionize chip design by automating many of the tedious tasks involved in the development process. Contestants have used these AI tools and others to generate Verilog code from natural language prompts, which they then implemented on the Efabless chipIgnite platform using the OpenLane open-source design flow. The contestants ranged from industry experts to people who had never before designed a chip. None of the winners had ever used the OpenLane Flow before.
Winning designs were selected by a panel of industry-respected judges who used pre-determined criteria, including design completeness, documentation, technical merit, and community interest. Efabless will now fabricate the three winning designs on its chipIgnite shuttle with the winners receiving packaged parts and evaluation boards – a value of $9,750. In addition, the other participants with qualifying designs will receive a free evaluation board and one of the winning AI-generated chips.
“I would like to congratulate the winners for creating such impressive designs in such a short amount of time,” said Mike Wishart, CEO of Efabless. “We were struck by the enthusiasm and collaborative participation of the community and the tremendous learning that was achieved by everyone involved. We look forward to our role in facilitating this innovation by providing a forum for learning and a platform to drive AI-assisted development from idea to silicon.”
Efabless will soon release information about the second AI-Generated Design Challenge. The challenge will take place over the summer, with tapeouts expected in September.
