Bringing Generative AI to Semiconductor and Electronics Design
Published Nov 15 2023 09:00 AM 10.2K Views
Microsoft

Generative AI promises to revolutionize how we live, work and play. This, as with prior technology innovations, have all been developed on increasingly complex, high-performance semiconductor platforms and systems. As such, it behooves us to utilize the transformative power of generative AI for accelerating innovation, improving quality, and enhancing safety of semiconductors.

 

Synopsys is an industry leader in Electronic Design Automation (EDA) providing foundational intellectual property (IP) and software design tools which are used across the industry to design and manufacture semiconductor chips. They also pioneered Synopsys.ai™, the industry’s first AI-driven EDA suite. A multi-year collaboration between Synopsys and Microsoft brought hyperscale cloud and AI technology to EDA, including the collaboration to enhance chip development Synopsys Cloud leveraging Microsoft Azure. With the advent of AI, we have been collaborating closely across our engineering and AI teams to bring this technology to market, in the form of Synopsys.ai Copilot.

 

This blog details our journey across the companies to announce the Synopsys.ai Copilot capability and highlights the possibilities we see for AI technology to be applied towards semiconductors and microelectronics development.

 

mujtabahamid_0-1699987559470.jpeg

 

Generative AI for HW Design

Software product development has transformed from a linear, siloed process to a continuous DevOps cycle. This was enabled by the cloud, and created a continuous loop for software, leading to faster product iteration, better quality through earlier collaboration, and lower cost of development and operation. AI, introduced with the GitHub Copilot, has gained rapid adoption and was powering over 46% of new code as of earlier this year.

 

Hardware systems design for semiconductors is starting to evolve along a similar trend, and a software centric design, verification and validation tool flow is well established in the industry. AI has the potential to transform this flow, from design authoring and register-transfer level (RTL), through the various steps in design verification and implementation, leading towards systems automation.

 

Hackathon and Joint Innovation

With this broad area of potential opportunity to apply AI, Microsoft and Synopsys conducted joint hackathons which brought together Azure Open AI implementation experts with Synopsys engineering, to get hands-on experience in integrating general AI frameworks from Microsoft to EDA specific tool flows. This included experiments with using custom data to enhance existing models, tune them, and apply prompt engineering to achieve desired results. Hackathons are an established tradition at Microsoft, where colleagues across the company are encouraged to team up to ideate and incubate new solutions such as the one in question here. The hackathon also showcased the innovation that is possible in a short time, with the right development tools, frameworks, and access to the right expertise.

 

Microsoft’s Silicon Design Team

To deliver the most comprehensive cloud solutions, Microsoft takes an approach of optimizing the entire stack, from energy draw at facilities, software design and server and silicon design.

 

Like many teams at Microsoft, the silicon team has experimented with  AI technology to run proof of concepts that have helped shape our approach to the use of AI for the industry. Understandably, there is a lot of excitement and optimism around it. "We're already seeing AI accelerate the design process by flattening learning curves, eliminating tedious work,  and upleveling expertise, which expands skillsets," says Erik Berg, Principal Engineer in the verification and validation team at Microsoft. "This technology not only helps us address issues early on, but also reduces the cost of fixing issues at a later stage. Incorporating AI into the silicon design and verification process has demonstrated remarkable results and added significant value to our work at Microsoft."

 

Microsoft silicon teams are already using early versions of Synopsys.ai Copilot to support our own engineering needs, and providing feedback, completing the virtuous product development cycle. 

 

 

Microsoft has a long history of deep partnership with the semiconductor and electronics industry. The collaboration with Synopsys on Synopsys.ai Copilot promises to be a transformative moment for the industry, sparking a new wave of innovation acceleration for a foundational part of the computing ecosystem. Stop by the Synopsys booth at Microsoft Ignite 2023 to learn more about Synopsys.ai CoPilot and the Microsoft ISV Partnerships teams to discuss opportunities to accelerate your company’s AI journey.

Co-Authors
Version history
Last update:
‎Nov 14 2023 11:02 AM
Updated by: