Google claims to have developed artificial intelligence software that can design computer chips faster than humans.
The tech giant said in an article in the journal Nature on Wednesday that a chip that would take people months could be created by its new AI in less than six hours.
AI has already been used to develop the latest iteration of Google’s tensor processor chips, which are used to perform AI-related tasks, Google said.
“Our method was used in the production to design the next generation of Google TPU,”
In other words, Google uses AI to design chips that can be used to create even more complex AI systems.
In particular, Google’s new AI can create a “floor plan” of the chip. This essentially involves graphics, where components such as processors, GPUs and memory are placed on top of the silicon matrix relative to each other – their positioning on these miniature boards is important as it affects the chip’s power consumption and processing speed.
People take months to optimally design these plans, but Google’s deep reinforcement system – an algorithm that is trained to take certain actions to maximize its chances of winning a prize – can do so with relatively little effort.
Such systems can also defeat people in complex games such as Go and chess. In these scenarios, algorithms are trained to move pieces that increase their chances of winning the game, but in the chip scenario, AI is trained to find the best combination of components to make it as computationally efficient as possible. The AI system was powered by 10,000 chip plans to “learn” what works and what doesn’t.
While human chip designers typically distribute components in clean lines, Google’s AI uses a more dispersed approach to designing its chips. This is not the first time an AI system has been tortured after learning how to perform a task behind human data. DeepMind’s famous “AlphaGo” AI made an extremely unconventional move against world champion Go Go Lee Sedol in 2016, which stunned Go players around the world.
Google engineers note in the article that the breakthrough could have “major consequences” for the semiconductor sector.
Facebook’s chief artificial intelligence scientist Ian LeKun hailed the study as “very good work” on Twitter, adding “this is exactly the type of setting in which RL shines.”
The breakthrough was hailed as an “important achievement”, which will be “a huge help to speed up the supply chain” in a Wednesday editorial in Nature.
However, the magazine says that “technical expertise must be shared widely to ensure that the companies'” ecosystem “becomes truly global.” He went on to stress that “the industry needs to make sure that time-saving techniques do not repel people with the necessary basic skills”.