We're only in year two of a "massive ten-year cycle" of rapid AI advancements and infrastructure build-out, AMD CEO Lisa Su said Wednesday at the Axios AI+ Summit in Washington, D.C.
Why it matters: The enormous investments being poured into data centers, chip development, and AI models are reorienting the U.S. economy, pumping up Wall Street valuations and gobbling up huge amounts of energy.
- Su expects a lot more of all of that, and thinks we can't even envision some advancements AI will enable five years from now.
Driving the news: Su, who has helmed the U.S. chipmaker for more than a decade, told Axios' Ina Fried her company's resurgence was made possible by focusing on the "bleeding-edge, highest-performing chips out there," which are in particularly high demand thanks to the AI boom.
- She called AI the "most transformative" technology of her long career in tech, and possibly any of our lifetimes.
- And she said the U.S had to ""run fast, and run even faster" to maintain its leadership in AI technology.
What she's saying: Su praised the Trump administration's focus on tech leadership and called its AI action plan a "good blueprint."
- She said her advice to political leaders on how to ensure the U.S. maintains its leadership in AI would be to encourage manufacturing in the U.S., make it easier to build data centers and ensure there's sufficient energy to power them.
Friction point: Su said she disagreed with Anthropic CEO Dario Amodei, who spoke immediately before her, that the U.S. should ban chip exports to China.
- "Our most advanced chips are export controlled, and they should be export controlled, but there is also an opportunity for us to get an AI stack that is based on American technology out into the world, and I think that is a good thing," she said.
- Along with Nvidia, AMD has agreed to pay the U.S. government 15% of its revenue from advanced chip sales to China.
The big picture: Su noted that market for AI data center accelerators was expected to balloon to $500 billion over the next three to four years.
- "Not too long ago, the entire semiconductor industry was $500 billion. So you can see the rate and pace of acceleration," she said.
What to watch: Su said she believes that within five years, AI will be solving problems we previously thought were impossible.
- One of her hopes is that five years from now, people will be able to get Mayo Clinic-level health care in rural America because of the medical discoveries and diagnostics enabled by AI.
