From Silicon To Software


The Ins and Outs of AI Chip Design

ai chip design flow
By Editorial Team

What are the ethics involved in a machine that makes decisions? If it’s deciding the next move on a chess board, maybe that’s a cool parlor trick. But, if your autonomous vehicle is making instantaneous decisions on traffic hazards, you understand what’s at stake. We are in a boom of AI-enabled applications that are quickly moving from simple AI acceleration to complex cognitive capabilities, able to process volumes of information that outstrip human capacity. If you want to peer into the world of what’s next in AI, check out our hot-off-the-presses post: AI’s Next Act: 5 Key Applications and Trends to Watch for in 2022. The possibilities for a smart everything world are quite literally mind boggling.

But what’s behind it? We’ve been talking about the ins and outs of AI chip design for a while now. As a refresher, we’ve compiled a retrospective of our best articles from 2021 on the subject to help you stay informed.

Here are our top five articles from our technical experts:

AI and AI Chip Design – A New “Chicken and Egg” Riddle

In this post, Mike Gianfagna posits the question, “Which came first, AI or the chips that accelerate AI?” In fact, applying AI to design AI chips is a game changer, and Mike provides a great overview of the industry buzz on this topic.

Beyond Machine Learning and Homomorphic Encryption: Why Advanced Verification Is Key to Fueling a New Era of AI Chips

Kiran Vittal shows us that it’s not just semiconductor companies designing today’s AI chips. The explosion of AI chips has changed the silicon engineering landscape. With these changes come challenges and opportunities to expand AI chips across applications. Learn why advanced datapath verification is critical to take us from an age of AI acceleration to cognitive systems.

The Real Impact of AI on Chip Design Is Autonomous Insight

Gianfagna is back again, discussing how complex chips requiring autonomous design can give designers visibility into an infinite number of possibilities and deliver autonomous insight for a new era.

Why Now Is the Time to Create an AI Strategy for Chip Design

Stelios Diamantidis explores how AI can take the guesswork out of chip design to achieve power, performance, and area targets faster, freeing engineering expertise up for more value-added work.

Synopsys and IBM Research: Driving Real Progress in Large-Scale AI Silicon and Implementing a Hybrid Cloud Model for Chip Design

Arun Venkatachar gives an overview of the Synopsys partnership with the IBM Research AI Center and how we are collaborating to improve AI performance by 1000x over the next decade.

Continue to stay on top of what’s happening in AI chip design and AI-enabled applications by following our “From Silicon to Software” blog.