AI Powers โNew approaches to Chipโข Design, Possibly Overcoming Physical Limits
SAN FRANCISCO, October 26, 2023 – โArtificial intelligence is being deployed to fundamentally rethink semiconductor design, offering a potentialโค path to circumvent the physical limitations increasingly hindering traditional chip development. Breakingviews reportsโค that companies are leveraging AI to explore novel chip architectures and materials, moving beyond conventional silicon-based designs.
For decades, Moore’s Law – the โฃobservation that the number ofโค transistors โขon a microchip doubles approximately every two โyears – has driven technological progress. Though, as transistors shrink to near-atomic levels, quantum effects and heat dissipation pose significant challenges. AI โoffers โฃa way to optimize designs at a level of complexity beyond human capability, potentially extending the โคlifespan of chip innovation.
Theโค approach involves using algorithms to explore โa vast design space, identifying configurations that would be impractical for engineers toโค conceive manually. This includesโฃ experimenting with new materials beyond silicon, like gallium nitride and silicon carbide,โ and unconventional chip layouts.โ “We’reโ essentially reinventingโ the silicon wheel, but with a logic engine that can explore possibilities we never could โbefore,” said one industry expert familiar with the technology, as reported by Breakingviews.
Several companies are already pursuing this strategy. Synopsys, a leading electronic designโ automation (EDA) software provider, is integrating AI into its tools to automate chip layout and verification.Startups like Tenstorrent,founded by former Google AI researcher Jim Keller,are designing chips specifically optimized for AI workloads,utilizing AI in the designโข process itself.
The potential benefits extend โขbeyond simply maintaining Moore’s Law. AI-driven design could leadโ to chips that are more energy-efficient,faster,and better suited for specialized โapplications like artificial intelligence,autonomous vehicles,and high-performance computing.
Robyn Mak, a Reuters Breakingviews journalist covering technology,โ has followed the development of AI in chip design since 2013,โ noting the increasing investment and sophistication of โthese tools. The โshift represents a significant departure from the established EDA industry andโ could reshape the future of semiconductor manufacturing.