AMD CEO Dr. Lisa Su says she will see AI 'serving to us design the next-generation GPU a lot quicker and extra reliably'
AI appreciation day might need formally been on 16 July, but it surely appears extra prefer it was the final 24 hours, what with Trump’s AI motion plan and the ‘Successful the AI Race’ summit. Along with the president, this latter occasion was additionally attended by each AMD CEO Dr. Lisa Su and Nvidia CEO Jensen Huang. And amidst some anticipated normal ‘reward AI’ fluff, Dr. Su stated one thing attention-grabbing: Subsequent-generation AMD GPUs may be made with the assistance of AI.
This was in response to being requested whether or not any R&D is ongoing at AMD or different locations that includes getting AI to give you new methods of producing—”completely different approaches that we might not essentially perceive.”
“We take into consideration how AI will design future chips, and it’ll design items of it” stated Dr. Su. “However there’s nonetheless a creativity of bringing all of it collectively that I feel people are nonetheless completely on the centre of… So I do not essentially see the AI designing our next-generation GPU, however I do see it serving to us design the next-generation GPU a lot quicker and extra reliably.”
In fact, that is all within the context of AI accelerators, but when AI is used to assist design these, we could be fairly assured it might be used to design the GPUs on the coronary heart of a few of the next-gen finest graphics playing cards. And if AMD’s doing it, we will guess Nvidia’s doing it, too.
In truth, we all know as a lot, as a result of Nvidia already makes use of AI to hurry up its chip design. That primarily seems to be to be within the type of a chatbot to assist engineers and others concerned discover paperwork and so forth, though its makes use of are presumably ever-expanding.
The AMD CEO appears eager on emphasising the significance of AI for the US, simply as Huang and Trump are, arguing that education must be gearing future generations in direction of producing the perfect AI expertise. She additionally has a constructive outlook on the end-goal of AI:
“You’ll be able to really take AI and make science higher, you’ll be able to take AI and make medication higher, you’ll be able to take AI and make manufacturing higher. You’ll be able to take AI and make each facet of your small business higher, and so in my thoughts, 10 years from now, we might prefer to consider that we’re actually leveraging it to unravel a few of the world’s most necessary issues.”
Properly, creating higher and cheaper graphics playing cards for gaming won’t be fixing one of many world’s most necessary issues—sorry, fellow avid gamers—but it surely’ll be a pleasant facet impact, if it really pans out as she describes.
With the sum of money AI firms are borrowing and spending, it would be good to see a few of that trickle down into avid gamers’ expectant palms. Now that is the sort of trickle-down economics I can get behind.
Greatest CPU for gaming 2025