Google is using AI to design AI processors much faster than humans can

(Image credit: Pixabay)

To one extent or another artificial intelligence is practically everywhere these days, from games to image upscaling to smartphone "personal assistants." More than ever, researchers are pouring a ton of time, money, and effort into AI designs. At Google, AI algorithms are even being used to design AI chips.

This is not a complete design of silicon that Google is dealing with, but a subset of chip design known as placement optimization. This is a time-consuming task for humans. As explained by IEEE Spectrum (via LinusTechTips), this involves placing blocks of logic and memory (or clusters of those blocks) in strategic areas to make the most of the available real estate, both for performance and power efficiency.

It might take a team of engineers several weeks to map out the ideal placement because it's a complex task with a ton of variables. In stark contrast, Google's neural network can produce a better design for a Tensor processing unit in less than 24 hours. This is similar in concept to the Tensor cores that Nvidia uses in its Turing-based GeForce RTX graphics cards, just with different goals in mind.

That's interesting in and of itself, but equally so is the type of AI Google is using. Rather than leverage a deep learning model, which requires training AI on a large set of data, Google is using a "reinforcement learning" system. The short explanation is RL models learn by doing.

There is a reward system involved, so RL models proceed in the right direction. In this case, the reward is a combination of power reduction, improvements in performance, and area reduction. I'm simplifying a bit, but basically, the more designs Google's AI does, the better it becomes at the task at hand (making AI chips).

"We believe that it is AI itself that will provide the means to shorten the chip design cycle, creating a symbiotic relationship between hardware and AI, with each fueling advances in the other," Google's researchers explain. If this works out for Google, it seems inevitable AMD, Intel and Nvidia will eventually try the same approach, too.

You check out the technical details in a paper posted to Arxiv.

Thanks, LinusTechTips.

TOPICS
Paul Lilly

Paul has been playing PC games and raking his knuckles on computer hardware since the Commodore 64. He does not have any tattoos, but thinks it would be cool to get one that reads LOAD"*",8,1. In his off time, he rides motorcycles and wrestles alligators (only one of those is true).

Latest in Processors
A chip being held up in an Intel fab
Intel is reportedly 'working to finalize commitments from Nvidia' as a foundry partner, suggesting gaming potential for the 18A node
AMD Strix Point APU chip, held in a hand, with the reflected light showing the various processing blocks in the chip die
AMD's next-gen 'Gorgon Point' APU outted and seemingly sticks with RDNA 3.5 graphics which is disappointing for handheld gaming PCs if accurate
Nvidia CEO Jensen Huang delivering pancakes and sausages to pre-GTC show hosts and guests, wearing an apron
'There might be a party. I wasn't invited,' says Jensen Huang of the rumoured TSMC proposal to join forces and run Intel's chip fabs
Nvidia Feynman GPU
While we despair of RTX 50-series supplies and wait on next-gen Rubin, Nvidia reveals its next-next GPU architecture will be known as Feynman and is due in 2028
Nvidia Vera CPU
Nvidia reveals Vera, a new CPU with 'custom' cores which could be very exciting for its upcoming premium PC processor
Machinery tools and equipment,Rolls of galvanized steel for production metal pipes and tubes for industrial ventilation systems in factory.
New super-thin '2D' metal sheets could enable ultra-low power chips and can you guess how they're made? Yup, by squishing stuff really hard
Latest in News
A female Zoi making two hearts with her fingers.
Following 24 hours of Denuvo-based backlash, Inzoi is taking a surprising step and removing it entirely: 'We want to sincerely apologise for not aligning more closely with player expectations'
An image of a Helldiver from Helldivers 2 shooting at a red dragon from Dungeons & Dragons.
'Ok, so dragon builds are a thing now': galaxy-brained Helldivers 2 player incinerates a bile titan with a hover pack and a flamethrower
An ancient, angry stone mech from No Man's Sky's new Relics update
No Man’s Sky lets you unearth ancient, angry mechs in the astro-archaeology filled Relics update
Assassin's Creed Shadows promo image
Ubisoft scores a legendary ratio against Elon Musk on his own platform—which hopefully marks a final end to all the Assassin's Creed Shadows' culture war nonsense
Tzarina Katarin Bokha, the Ice Queen of Kislev
Total War: Warhammer 3 rolls out a cool Kislev overhaul, changes befitting Tzeench’s magic, new projectile units and creakier skeletal horses
An image of a golden first place award from Geoguessr
'We're actually getting GeoGuessr on Steam before GTA 6': the Google Street View puzzler arrives on Valve's platform this April