Dec 15, 2023
Lisa Su Saved AMD. Now She Wants Nvidia's AI Crown
From a conference room atop AMD’s headquarters in Santa Clara, California, a
From a conference room atop AMD's headquarters in Santa Clara, California, a stretch of highway 101 running outside, Lisa Su presides over a company older than the term "Silicon Valley." Down the road is a link to the company's past, an old foundry in Sunnyvale where AMD used to press its chips. But from her window she can see a recent milestone in the company's fast-evolving present: the offices of arch-nemesis Intel, whose market capitalization ($120.3 billion) AMD's now eclipses ($153.5 billion).
It wasn't always this way. In 2014, when Su, now 53, took up the CEO reins at AMD (Advanced Micro Devices), the chipmaker was foundering. The company had laid off around a quarter of its staff and its share price hovered around $2. Patrick Moorhead, a former AMD exec, remembers it as "deader than dead." Then Intel began to stumble, dragged down by manufacturing delays and Apple's decision not to use its chips in iPhones. Nimble, with a tactician's eye, Su was able to capitalize on her rival's missteps, inking deals with laptop makers such as Lenovo and gaming giant Sony, plus Google and Amazon, whose massive data centers generated $6 billion of the chipmaker's sales last year.
"If you look out five years, you will see AI in every single product at AMD, and it will be the largest growth driver."
At $63 billion, Intel's annual revenue still dwarfs AMD's $23.6 billion. But wresting away coveted server chip market share from its Silicon Valley neighbor, as well as scooping up the semiconductor company Xilinx, has spiked AMD's stock nearly 30-fold in the nine years since Su took over. Now, with the mainstreaming of artificial intelligence stoking demand for the silicon brains behind machine learning, she's facing a legacy-defining opportunity and a daunting challenge: Can AMD produce a chip powerful enough to break Nvidia's near-monopoly on the processors that undergird the coming wave of generative AI technology? "If you look out five years," she says, "you will see AI in every single product at AMD, and it will be the largest growth driver."
Su has been overclocking AMD for the last nine years, much like a gamer who pushes a processor to perform beyond its manufacturer-specified limits. Unlike many tech executives, she's a world-class researcher, with a Ph.D. in electrical engineering from MIT. Her unique combination of technical genius, people skills and business savvy has made her among the highest-paid S&P 500 CEOs for the last several years (total 2022 compensation: $30.2 million). Overall, she has amassed a $740 million fortune (largely in AMD stock), landing her in 34th place on our annual ranking of America's richest businesswomen. "Talk about leaning in and just killing it," marvels Panos Panay, Microsoft's chief product officer, who first met Su in 2014 as she began AMD's turnaround.
Unlike Intel, though, whose revenue has seen a 12% decrease to $63.1 billion over three years, Nvidia appears at the top of its game. Beyond rendering stunning imagery in games like Cyberpunk 2077, its GPUs (graphics processing units) have become the engine of choice for artificial intelligence companies such as OpenAI, whose ChatGPT chatbot has delighted and disturbed the public by answering questions and commands with surprisingly detailed human-sounding responses.
These so-called large language models are really just stunning parlor tricks, but they’re the opening act for an AI transformation that big shots like Bill Gates say will be as significant as the dawn of the internet. Already there is huge demand for the GPUs that power them, and at least one research firm foresees a $400 billion bonanza within the next decade for the companies that make them. But right now there's really just one. "AI equals Nvidia," says Glenn O’Donnell, a Forrester analyst. "That's pretty well-entrenched, and AMD has to really step up its game to overcome that."
Meanwhile, the specter of Intel still looms across Highway 101, even as the O.G. of PCs has faced further manufacturing delays, chip defects and leadership changes. "There are many great things about AMD, but the bad thing is that we have two world-class competitors," says AMD exec Forrest Norrod, who helped Dell build its approximately $10 billion (2014 revenue) data center business in part on AMD chips and adds that the company never assumes its main rival will let problems linger. "We will always assume that Intel will fix it."
Su is taking on Nvidia in the AI wars. Her secret weapon: the Instinct MI300 (shown here), a chip that melds traditional CPUs with GPU processors that are often used in gaming.
WHEN SU WAS promoted into AMD's top job in 2014, analysts were calling the company "uninvestable," with $2.2 billion in debt. Some of its prized assets were already being sold for parts. Its fabrication plant where chips are baked ("fabs," in industry-speak) was spun off in 2009—a blow to AMD cofounder Jerry Sanders’ infamous boast that "real men have fabs." It even had to sell and lease back its corporate campus, in Austin, Texas—Su's current base—in 2013.
More troubling, AMD was struggling to execute. It couldn't hit product deadlines, and Intel dominated all but the bargain end of the laptop market with Nvidia, Qualcomm and Samsung carving up the new smartphone business. "Our technology wasn't competitive at the time," Su admits.
AMD hadn't always been such a headache for investors. Sanders broke into the microprocessor business making chips for IBM in the early 1980s, but things started to change in the late ’90s and early 2000s. AMD, which had been a perennial second-rater, began generating record profits by building its own processors that beat Intel's on speed.
By 2014, those glory days were long gone. As were about a quarter of AMD's staff, sacked by Su's predecessor, Rory Read (Sanders stepped down as CEO in 2002). AMD had once commanded about a quarter of the now $24 billion server chip market, but its share dwindled to 2% in 2014. On her second day as CEO, Su stepped up to the microphone during an all-hands call with a message for AMD's demoralized employees: "I believe that we can build the best," she remembers telling her staff. "You might think that was obvious, but it wasn't to the company at the time."
That rallying cry was also an edict and step one in her three-pronged plan to fix AMD: Create great products, deepen customer trust and simplify the company. "Three things, just to keep it simple," she says. "Because if it's five or ten, it's hard."
Su refocused her engineers on building Intel-beating chips, but it can take chip designers years to draw up a viable final blueprint. AMD's share of the server market fell even further, to half a percent, while researchers toiled in the lab. "At the time, the company wasn't doing well, but holy cow, they were working on the most exciting design in the industry," she says. "Engineers are motivated by products, and I like to keep that front and center."
Her decision to prioritize a new chip architecture called Zen paid off when it finally launched in 2017. "It was really good," she says with palpable pride, adding that Zen could compute more than 50% faster than the company's previous designs. More importantly, it signaled to the industry that AMD had turned a corner. By Zen's third generation, released in 2020, it was the market leader in terms of speed. Zen architecture now underpins all of AMD's processors.
With her team shepherding a new generation of chips, Su hit the road to hard-sell them to jaded data center clients. She had already spent years building relationships even when AMD had no chips to sell, once driving more than four hours through a Texas ice storm to charm Antonio Neri, now CEO of Hewlett Packard Enterprise. "I was, let's say, disenchanted by AMD's prior generation," Neri says. "She showed me she had the conviction of what needed to be done."
A big part of Su's strategy was inking new deals with the tech giants, which needed oodles of CPUs to power their exploding cloud businesses. "For us, there are really three microprocessor partners. We have Nvidia, Intel, AMD," says Thomas Kurian, CEO of Google Cloud. "AMD, when I joined, was not really a significant part of our ecosystem—at all. And it is a credit to Lisa that they are a very important partner for us now."
Last February, when AMD's market cap first surpassed Intel's, company cofounder Sanders, now 86, was ecstatic. "I called everybody I know!" he says. "I was delirious. I’m only sorry that Andy Grove isn't around so I could say ‘gotcha!’ " (Grove, Intel's legendary former CEO, died in 2016.)
SU
, THE DAUGHTER of a mathematician and a bookkeeper turned entrepreneur, was born in Tainan, Taiwan, in 1969, the same year Sanders founded AMD. Her family immigrated to New York City when she was 3. She chose electrical engineering at the Massachusetts Institute of Technology because it seemed to be the most difficult major. For someone so technically talented, she was also good with people, playing peacekeeper when disagreements flared between fellow students, says Hank Smith, who ran MIT's nanostructures lab at the time.
AMD cofounder Jerry Sanders (shown here in 1979) was an early Silicon Valley swashbuckling salesman. "People thought I was all hat and no cattle," he tells Forbes now. But he argues AMD wouldn't have lasted so long if that were true. "We were just trying to get attention. We were in the shadow of Intel."
Su laughs upon hearing herself described as a people person. "Well, that's compared to other MIT people," she jokes. "I don't think anyone would say I’m an extrovert, but communication is a huge part of my job."
After a brief stint at Texas Instruments, in 1995 she was hired as a staff researcher at IBM, where she helped design chips that run 20% faster by using semiconductors with copper circuitry instead of the traditional aluminum. Higher-ups quickly spotted her talent: In 1999, a year after the launch of the copper technology, IBM's then-CEO, Lou Gerstner, tapped her to be his technical assistant. In his first interview in 20 years, Gerstner tells Forbes he initially worried Su was too junior for the job, but his doubts were quickly quelled. "She proved to be one of the most outstanding employees who worked in my office. Lisa doesn't follow normal patterns—she's been blowing them up her entire career."
That appointment gave Su a front-row seat for a corporate turnaround that is now a classic business school case study—a rejuvenation ignited, in part, by leaning on the company's scale and creating a culture devoted to customers. Gerstner grew the stalled IBM's market cap nearly sixfold in his almost nine-year run. Su also got a taste for dealmaking, helping IBM sign a joint deal with Sony and Toshiba in 2001 to put its chips in Sony's PlayStation 3.
Early on, she sometimes worried she wasn't qualified to sit at tables packed with business heavyweights, but Su soon realized that her hard-learned technical acumen gave her an edge over executive-track types. "I saw that MIT Ph.D.'s worked for Harvard MBAs, and the truth is that made absolutely no sense to me," she said in a 2017 graduation speech at her alma mater, whose new nanotech laboratory now carries her name.
In late 2011, Nick Donofrio, at the time an AMD board member whom Su had met at IBM, called up his old protégée, who by then was a senior vice president of Free-scale, the Austin-based chipmaker that is now part of NXP Semiconductors. The two met for dinner, and over a bottle of Brunello, he made his pitch: an opportunity not just to chase incremental improvements, but to reinvent and innovate—with the ground cover to actually do it.
A few days later, Su accepted a role as senior vice president of AMD's global business units. Two years after she started, she was running the entire company—making her the first female CEO of a major semiconductor company.
"I would walk into rooms where there were, like, 25 people, and I might have been the only woman," she recalls of her early engineering days. "Where I have a lot of passion is young women engineers—keeping them in engineering."
"AMD, when I joined, was not really a significant part of our ecosystem—at all. And it is a credit to Lisa that they are a very important partner for us now."
When Su first took over, she flew to Beverly Hills and personally asked AMD's Sanders to speak to her team. Sanders says he was touched by Su's offer but declined. "It's not my team now. It's your team," he recalls telling her. Ever the salesman, though, he also made a counteroffer: He’d make the visit once the company had hit two years of profitability. In 2019, coinciding with the company's 50th anniversary, Sanders made good on that promise.
Semiconductor hotshot Mark Papermaster, who led the iPhone and iPod engineering teams at Apple and who had joined AMD around the same time as Su, has been a keen observer of the company's remarkable comeback under her leadership. At Apple, Papermaster had worked for another skilled turnaround artist: cofounder Steve Jobs, who saved the company from catastrophe and set it on the path that would make it the world's most valuable company. "What Lisa had, in a number of ways, was an even more difficult task," Papermaster says. "When you’re not a founder, you have to establish your own credibility and your own vision, and bring the entire company, your customers and your investors with you."
Su's success at AMD has made her an inspiration for young engineers and a hero to investors. It has also made her into a meme: A few years ago, 8-bit animations in which Su uses AMD's Ryzen chips to morph into a superhero or shoot lasers from her eyes went viral on Twitter. A figurine of her clad in orange-and-red armor, helmet at her side, features prominently on her office shelf, a gift from a fan at the E3 gaming conference. "That's probably one of the funnier moments in my career," says Su, who, while a keen Twitter and Reddit user, isn't "big on memes. It's not my thing."
NOW THAT SU has renewed and energized AMD, she's focused on ensuring its future in a highly competitive market. While she diligently rebuilt its business, Nvidia cofounder and CEO Jensen Huang was hard at work making his company the go-to vendor for artificial intelligence computing power.
Huang, who is a distant relative of Su's, sees a gold mine in selling the chips to buttress AI tools like ChatGPT. Demand has already catapulted Nvidia's share price to near all-time highs with a forward P/E of around 64x—nearly double AMD's. "It's why investors are looking at AMD: because they want the poor man's Nvidia," says Stacy Rasgon, an analyst at Bernstein. "Maybe the market is so big they don't need to be competitive."
Su is in a good position to take a run at the AI chip market. But she knows well how quickly turnarounds can become downfalls.
But Su intends it to be. And she hopes to take on Nvidia's AI-centric H100 GPUs by betting on annual chip upgrades meant to burnish AMD's position. Under her leadership, R&D spending has risen nearly fourfold, to $5 billion—almost as much as AMD's entire revenue when she took over.
A new supercomputer at Tennessee's Oak Ridge National Laboratory—the fastest in the world when completed in 2022—is Su's passion project. The groundbreaking machine was built to have the processing power of at least a quintillion calculations per second and is a showcase for AMD's AI chips. She's throwing a curveball as well: The MI300 chip, which fuses CPUs with GPUs in a bid to counter Nvidia's new superchip, will ship later this year.
She has also been maneuvering against Nvidia with acquisitions, such as her $48.8 billion takeover in 2022 of Xilinx, a company that makes programmable processors that help speed up tasks like video compression. As part of the deal, Victor Peng, Xilinx's former CEO, became AMD's president and leader of AI strategy.
Beyond Nvidia lurk other emerging threats: Some of AMD's customers have begun doing chip development of their own—a move designed to mitigate their dependence on the semiconductor giants. Amazon, for example, designed a server chip in 2018 for its AWS business. Google has spent nearly a decade developing its own AI chips, dubbed Tensor Processing Units, to help "read" the names of the signs captured by its roving Street View cameras and provide the horsepower behind the company's Bard chatbot. Even Meta has plans to build its own AI hardware.
Su shrugs off concerns that her customers could someday be competitors. "It's natural," she says, for companies to want to build their own components as they look for efficiencies in their operations. But she thinks they can do only so much without the technical expertise AMD has built over the decades. "I think it's unlikely that any of our customers are going to replicate that entire ecosystem."
Su is in a good position to take a run at the AI chip market. But she knows well how quickly turnarounds can become downfalls. There's more work to be done to ensure AMD endures: "I think there's another phase for AMD. We had to prove that we were a good company. I think we’ve done that. Proving, again, that you’re great, and that you have a lasting legacy of what you’re contributing to the world, those are interesting problems for me."
WHEN SU WAS SU , THE DAUGHTER NOW THAT SU MORE FROM FORBES