Synopsys Implements Microsoft Copilot to aid in Chip Designing

Microsoft has primarily marketed its "Copilot" as a tool for completing computer code lines or efficiently summarizing a cluttered email inbox. Synopsys announced on Nov 15 that it has collaborated with Microsoft to develop a custom Copilot for facilitating computer chip design. Designing chips is an extremely challenging undertaking in the technology industry due to the need for precise arrangement of billions of transistors, which are minuscule on-off switches, on a small piece of silicon that is only a few centimeters wide. Creating a chip usually requires an investment of hundreds of millions of dollars and a few years of work, even with a large team of engineers.

On Nov 16, Synopsys, the leading software maker for that process, announced that it utilized Microsoft's Azure OpenAI system to develop its own Copilot for integration with its tools. In the initial phases of creating a complex chip, engineers use a language that is similar to software programming code to outline how the chip should function.

Synopsys utilized its extensive collection of data amassed over several decades in business to train the system. According to Shankar Krishnamoorthy, the general manager of the design automation group at Synopsys, the main goal was to ensure the accuracy of the system.

According to Krishnamoorthy, even if an AI bot makes a 10 to 15 percent error when writing essays, Shakespeare, or poetry, the consumer would still be impressed and able to distinguish the difference. If your chip design work is not producing over 99.9 percent accuracy, you are essentially creating a bug in your chip, which could cost hundreds of millions of dollars.