There’s no doubt Google employs more than a few people with elite intelligence.
Those people, in turn, apparently have the inside track on artificial intelligence that the company claims can design computer chips at a much higher rate than humans can.
How much faster is the AI software?
The research, published in Nature on Wednesday, spells it out.
“The RL (reinforcement learning) agent becomes better and faster at floorplanning optimization as it places a greater number of chip netlists,” the researchers wrote in the study.
“We show that our method can generate chip floorplans that are comparable or superior to human experts in under six hours, whereas humans take months to produce acceptable floorplans for modern accelerators.”
Six hours compared with “months”? That could solve the global chip shortage in a big hurry.
Not to delve too deeply into the details, but Google’s form of AI already has helped develop the most recent version of Google’s tensor processing unit chips, used to run AI-related tasks, Google said.
“Our method has been used in production to design the next generation of Google TPU,” Google’s authors wrote.
To put it another way, Google is using AI to design chips that can be used to develop and implement more sophisticated AI systems.
Systems similar to this, a CNBC story reported, can “also defeat humans at complex games like Go and chess.”
In a Nature editorial, the AI accomplishment was called an “important achievement” that will “be a huge help in speeding up the supply chain,” but the editorial added, “the industry must make sure that the time-saving techniques do not drive away people with the necessary core skills.”