Nvidia’s AI-based chip design can help with better placement of transistors, lower costs
Nvidia Corp, the world’s leading designer of computer chips used to create artificial intelligence, presented new research on Monday that explains how artificial intelligence can be used to improve chip design.
In the chip design process, it is decided where to place tens of billions of small on-off switches called transistors on a piece of silicon to create working chips. The exact placement of these transistors has a big impact on the cost, speed and power consumption of the chip.
Chip design engineers use complex design software from companies such as Synopsys Inc and Cadence Design Systems Inc to help them optimize the placement of these transistors.
On Monday, Nvidia published a paper showing that it could use a combination of artificial intelligence techniques to find better ways to place large arrays of transistors. The paper aims to improve on Alphabet Inc’s Google’s 2021 paper, whose findings later sparked controversy.
The Nvidia study used efforts developed by researchers at the University of Texas using so-called reinforcement learning and added another layer of artificial intelligence on top of it to get even better results.
Bill Dally, chief scientist at Nvidia, said the work is important because improvements in chip manufacturing are slowing and the cost per transistor in new generations of chip manufacturing technology is now higher than in previous generations.
This is contrary to Intel Corp founder Gordon Moore’s famous prediction that chips would always get cheaper and faster.
“You don’t actually get any more economics from that scale,” Dally said. “We can’t keep going and provide more value to customers, we can’t get it from cheaper transistors. We have to get it by being smarter about the design.”
Read all the Latest Tech News here.
