The most popular and comprehensive Open Source ECM platform
Let us know more
Chip Design by AI: Google Algorithm Bootstraps First Generation of Self-Designed Chips
Google has turned over the design of their AI chips to AI algorithms. They are creating the first generation of AI technology that is being built by AI itself. The Google Brain Team reported in an article in nature magazine that AI algorithms are able to take requirements and create a design for new chips that exceed designs created by humans when judged by power consumption, performance, and physical chip size. AI design can be done in hours by AI algorithms compared to the many weeks and months that it would take humans to do the work manually.
Floorplanning is the task of designing the physical layout of a computer chip.
The Google Brain Team wrote that “we pose chip floorplanning as a reinforcement learning problem, and develop an edge-based graph convolutional neural network architecture capable of learning rich and transferable representations of the chip. As a result, our method utilizes past experience to become better and faster at solving new instances of the problem, allowing chip design to be performed by artificial agents with more experience than any human designer. Our method was used to design the next generation of Google’s artificial intelligence (AI) accelerators, and has the potential to save thousands of hours of human effort for each new generation. Finally, we believe that more powerful AI-designed hardware will fuel advances in AI, creating a symbiotic relationship between the two fields.”
After training on more than 10,000 previous AI chip designs, the algorithm used some unorthodox design patterns that surprised researchers who considered the techniques to be possible “breakthroughs ” in new design.
Ben Dickson, software engineer at TechTalks, wrote for Venturebeat that “I don’t see this as a story of ‘AI outsmarting people, ‘AI making AI smarter’ or AI developing ‘recursive self-improvement’ skills. Rather, it is a manifestation that people are finding ways to use AI as a prop to overcome their own cognitive limits and expand their capabilities. If there is a virtuous cycle, it is one in which AI and humans find better ways to work together.”
