WEBVTT 00:00:00.000 --> 00:00:02.060 Welcome back to the Deep Dive. You know, this 00:00:02.060 --> 00:00:04.419 is where we take that huge stack of professional 00:00:04.419 --> 00:00:07.059 source material, all the research, and really 00:00:07.059 --> 00:00:09.279 just boil it down for you. We want to hand you 00:00:09.279 --> 00:00:11.919 the keys to what actually matters. And today, 00:00:12.199 --> 00:00:14.880 wow, we are focusing on a leader whose impact 00:00:14.880 --> 00:00:18.399 is so profound, it's... Well, it's often called 00:00:18.399 --> 00:00:20.359 a corporate miracle. Oh, it really is. And when 00:00:20.359 --> 00:00:22.320 we talk about magnitude, we're not just talking 00:00:22.320 --> 00:00:24.260 about a company turnaround. We're talking about 00:00:24.260 --> 00:00:28.100 a fundamental structural shift in the entire 00:00:28.100 --> 00:00:30.859 global semiconductor industry. Yeah. Lisa Su, 00:00:30.960 --> 00:00:33.939 she's the president, chair and CEO of AMD. And 00:00:33.939 --> 00:00:35.640 she didn't just stabilize a struggling company. 00:00:35.700 --> 00:00:38.939 She pulled off one of the most incredible long 00:00:38.939 --> 00:00:41.219 shot turnarounds in modern history. Absolutely. 00:00:41.340 --> 00:00:43.859 We see so many executives who are brilliant at 00:00:43.859 --> 00:00:46.729 finance or maybe marketing. But what the sources 00:00:46.729 --> 00:00:49.929 show is that Sue's story is different. It's rooted 00:00:49.929 --> 00:00:53.530 in decades of deep foundational engineering expertise. 00:00:53.929 --> 00:00:56.189 She's one of those rare leaders who you feel 00:00:56.189 --> 00:00:58.750 genuinely understands how the electrons flow. 00:00:58.929 --> 00:01:01.070 Exactly. And the numbers, I mean, the numbers 00:01:01.070 --> 00:01:03.030 are just staggering. They really show the value 00:01:03.030 --> 00:01:05.430 she unlocked. When she took over as CEO back 00:01:05.430 --> 00:01:09.370 in 2014, AMD's market cap was, what, around $3 00:01:09.370 --> 00:01:12.930 billion? A worrying $3 billion. And today, under 00:01:12.930 --> 00:01:16.310 her leadership, It soared to over $200 billion. 00:01:16.590 --> 00:01:19.650 That's not just growth. That is a complete reinvention. 00:01:19.750 --> 00:01:21.769 Total Phoenix story. And it's more than just 00:01:21.769 --> 00:01:24.909 a resurrection. It's a true David versus Goliath 00:01:24.909 --> 00:01:28.329 victory. The single biggest metric of her success 00:01:28.329 --> 00:01:30.209 has to be this one. Oh, I know what you're going 00:01:30.209 --> 00:01:32.549 to say. During her tenure, for the very first 00:01:32.549 --> 00:01:36.189 time, A &E actually overtook its massive longtime 00:01:36.189 --> 00:01:40.000 rival, Intel. in market capitalization which 00:01:40.000 --> 00:01:42.120 just wasn't supposed to be possible intel was 00:01:42.120 --> 00:01:44.859 the permanent giant amd was the permanent underdog 00:01:44.859 --> 00:01:47.640 and suddenly they were winning the race it completely 00:01:47.640 --> 00:01:50.340 changed the game the entire competitive landscape 00:01:50.340 --> 00:01:53.140 for processors worldwide and that blend you mentioned 00:01:53.140 --> 00:01:55.200 the technical know -how and the strategic vision 00:01:55.200 --> 00:01:57.099 that's why she's not just a historical figure 00:01:57.099 --> 00:02:00.239 she's deeply deeply contemporary right and the 00:02:00.239 --> 00:02:02.439 sources point straight to that she was named 00:02:02.439 --> 00:02:05.099 one of the architects of ai for times person 00:02:05.099 --> 00:02:08.180 of the year in 2025. That title, Architect of 00:02:08.180 --> 00:02:10.979 AI, that says it all. It really does. It cements 00:02:10.979 --> 00:02:13.479 her role in shaping the current technological 00:02:13.479 --> 00:02:15.680 age we're living in. And it perfectly connects 00:02:15.680 --> 00:02:19.139 her earliest work in device physics directly 00:02:19.139 --> 00:02:21.219 to the biggest challenges we face today, where 00:02:21.219 --> 00:02:23.840 power and efficiency are everything in the AI 00:02:23.840 --> 00:02:26.939 race. So our mission for you in this deep dive 00:02:26.939 --> 00:02:30.800 is to really understand that connection. How 00:02:30.800 --> 00:02:33.099 did deep, fundamental engineering competence 00:02:33.099 --> 00:02:35.620 become the ultimate weapon in the boardroom? 00:02:36.080 --> 00:02:38.639 Okay, let's unpack this. Specifically, let's 00:02:38.639 --> 00:02:42.039 trace that competence. How did a background in 00:02:42.039 --> 00:02:46.120 experimental, really foundational engineering 00:02:46.120 --> 00:02:48.659 pave the way for such a massive industry shift? 00:02:48.900 --> 00:02:50.180 And to do that, you have to start right at the 00:02:50.180 --> 00:02:52.979 beginning. Right, with the foundations, her early 00:02:52.979 --> 00:02:55.560 life, and that relentless pursuit of engineering 00:02:55.560 --> 00:02:58.819 excellence. So Lisa Tsufang Su was born in Tainan, 00:02:58.840 --> 00:03:01.620 Taiwan, November 1969. And she wasn't there for 00:03:01.620 --> 00:03:04.030 long. She immigrated to the U .S. with her parents 00:03:04.030 --> 00:03:05.969 when she was just three years old. Settled in 00:03:05.969 --> 00:03:08.550 Queens, New York City. And, you know, that immigration 00:03:08.550 --> 00:03:11.090 context, it so often shapes the kind of drive 00:03:11.090 --> 00:03:13.490 and ambition we see in leaders like her. It really 00:03:13.490 --> 00:03:16.849 does. And the sources emphasize that the environment 00:03:16.849 --> 00:03:21.750 at home was, well, it was rigorous, but also 00:03:21.750 --> 00:03:24.469 incredibly supportive. You had this amazing blend 00:03:24.469 --> 00:03:29.699 of influences. Her father, Su Chui. was a statistician 00:03:29.699 --> 00:03:32.240 so you get that connotative analytical mindset 00:03:32.240 --> 00:03:35.039 right there absolutely and her mother sandy lowe 00:03:35.039 --> 00:03:37.780 started as an accountant but later became a successful 00:03:37.780 --> 00:03:39.840 entrepreneur so he's getting the business side 00:03:39.840 --> 00:03:42.379 the risk management from her mom that's a powerful 00:03:42.379 --> 00:03:45.199 combination it's a rare mix the precision of 00:03:45.199 --> 00:03:47.680 a statistician and the market sense of an entrepreneur 00:03:47.680 --> 00:03:50.419 all at the dinner table and the sources say both 00:03:50.419 --> 00:03:54.110 she and her brother were just pushed Pushed relentlessly 00:03:54.110 --> 00:03:56.610 to study math and science. She's told stories 00:03:56.610 --> 00:03:58.189 about her dad quizzing her on multiplication 00:03:58.189 --> 00:04:01.310 tables when she was just seven. Just seven. So 00:04:01.310 --> 00:04:02.830 those fundamentals were being drilled in from 00:04:02.830 --> 00:04:04.830 day one. Meanwhile, her mother is introducing 00:04:04.830 --> 00:04:07.110 her to business concepts. So she's getting this 00:04:07.110 --> 00:04:09.289 dual perspective from a really, really young 00:04:09.289 --> 00:04:11.909 age. And that blend, it's so telling when you 00:04:11.909 --> 00:04:14.789 look at her later career. It really is. It fostered 00:04:14.789 --> 00:04:17.490 this innate curiosity that seems to be the main 00:04:17.490 --> 00:04:19.750 driver for her more than just, you know, financial 00:04:19.750 --> 00:04:23.040 reward. The engineer's curiosity. That's what 00:04:23.040 --> 00:04:25.220 really stands out in the sources. Yeah, she said 00:04:25.220 --> 00:04:28.220 she wanted to be an engineer simply because she 00:04:28.220 --> 00:04:31.160 just had a great curiosity about how things worked. 00:04:31.420 --> 00:04:33.800 It wasn't about the job title or the status. 00:04:34.040 --> 00:04:37.220 It was about taking things apart to see the core 00:04:37.220 --> 00:04:39.579 mechanics. And we see that play out literally. 00:04:39.660 --> 00:04:41.620 There's this great anecdote about her at age 00:04:41.620 --> 00:04:44.259 10. Taking apart her brother's remote control 00:04:44.259 --> 00:04:47.259 cars. And fixing them. It's a perfect little 00:04:47.259 --> 00:04:50.620 metaphor for her entire career, right? Analyzing 00:04:50.620 --> 00:04:53.839 a complex, broken system and methodically finding 00:04:53.839 --> 00:04:56.199 the flaw. And then her first computer in junior 00:04:56.199 --> 00:04:58.860 high was an Apple II. The classic. So she got 00:04:58.860 --> 00:05:01.620 that early hands -on exposure. All that curiosity 00:05:01.620 --> 00:05:04.699 led her to a very competitive high school. She 00:05:04.699 --> 00:05:06.639 graduated from the Bronx High School of Science 00:05:06.639 --> 00:05:09.420 in 1986. Which is, I mean, it's known globally 00:05:09.420 --> 00:05:11.959 for its intense focus on STEM. It's a feeder 00:05:11.959 --> 00:05:14.240 school for top tech universities. And that's 00:05:14.240 --> 00:05:17.100 where she went next. a logical but very difficult 00:05:17.100 --> 00:05:20.639 jump to MIT in the fall of 86. She was initially 00:05:20.639 --> 00:05:22.980 thinking about either electrical engineering 00:05:22.980 --> 00:05:26.680 or computer science. But her final choice is 00:05:26.680 --> 00:05:29.740 so telling. It really reflects this pattern in 00:05:29.740 --> 00:05:32.720 her life of always picking the biggest challenge. 00:05:32.959 --> 00:05:35.199 She settled on electrical engineering, and her 00:05:35.199 --> 00:05:36.959 reason was that she thought it seemed like the 00:05:36.959 --> 00:05:40.220 most difficult major. Who does that? It shows 00:05:40.220 --> 00:05:42.740 this character trait, right? She gravitates toward 00:05:42.740 --> 00:05:45.000 the hard problems, the places with the most friction. 00:05:45.339 --> 00:05:47.259 Because that's where the real innovation and 00:05:47.259 --> 00:05:50.620 the real risk is. And her time at MIT, it wasn't 00:05:50.620 --> 00:05:53.439 just sitting in lectures. She got critical hands 00:05:53.439 --> 00:05:56.459 -on experience through the UROP program. The 00:05:56.459 --> 00:05:58.379 Undergraduate Research Opportunities Program. 00:05:58.660 --> 00:06:00.339 And this wasn't theory. She was an undergraduate 00:06:00.339 --> 00:06:02.579 research assistant, physically manufacturing 00:06:02.579 --> 00:06:05.879 test silicon wafers for grad students. Which 00:06:05.879 --> 00:06:07.899 might sound tedious, but for you, the listener, 00:06:07.980 --> 00:06:10.079 you have to understand how fundamental that is. 00:06:10.259 --> 00:06:12.300 Right. She wasn't just reading about semiconductors. 00:06:12.379 --> 00:06:15.199 She was in the clean room. doing the meticulous 00:06:15.199 --> 00:06:17.819 work of creating the very foundation for chips. 00:06:18.319 --> 00:06:20.879 That practical work, plus some summer jobs at 00:06:20.879 --> 00:06:23.639 analog devices, that's what cemented her interest 00:06:23.639 --> 00:06:26.980 in this super complex world of semiconductors. 00:06:27.100 --> 00:06:28.920 It pointed her directly toward that intersection 00:06:28.920 --> 00:06:31.060 of material science and electrical engineering. 00:06:31.199 --> 00:06:33.579 And that path leads directly to her doctoral 00:06:33.579 --> 00:06:37.180 work from 1990 to 94. And this is where she makes 00:06:37.180 --> 00:06:41.040 a really crucial early technical bet. Yes. Her 00:06:41.040 --> 00:06:43.339 Ph .D. work focused on a technology that was 00:06:43.339 --> 00:06:46.920 foundational, but at the time completely unproven. 00:06:47.230 --> 00:06:50.170 Under her advisors, Dimitri Antonidis and James 00:06:50.170 --> 00:06:52.230 Chung, her research is really on the cutting 00:06:52.230 --> 00:06:54.029 edge. She was one of the very first researchers 00:06:54.029 --> 00:06:56.990 anywhere to investigate silicon on insulator 00:06:56.990 --> 00:07:00.189 technology. Or SOI. Exactly, SOI. And that might 00:07:00.189 --> 00:07:02.149 sound like a bunch of jargon, but it's so important 00:07:02.149 --> 00:07:04.490 for understanding modern chip efficiency. It 00:07:04.490 --> 00:07:07.550 is. The sources explain that SOI was this unproven 00:07:07.550 --> 00:07:09.959 technique. The goal was simple but brilliant. 00:07:10.160 --> 00:07:12.819 Make transistors way more efficient and faster 00:07:12.819 --> 00:07:15.300 by isolating them. So a standard transistor is 00:07:15.300 --> 00:07:17.540 built right on the silicon wafer. Right. And 00:07:17.540 --> 00:07:19.500 when current flows through it, some of that electricity 00:07:19.500 --> 00:07:22.240 can leak away into the silicon underneath. It 00:07:22.240 --> 00:07:25.819 wastes power and it creates heat. But SOI introduces 00:07:25.819 --> 00:07:29.379 an insulating layer -like silicon dioxide between 00:07:29.379 --> 00:07:31.800 the transistor and that main silicon substrate. 00:07:32.300 --> 00:07:35.240 Think of it like a dam. It's a barrier. It stops 00:07:35.240 --> 00:07:37.800 that leakage, which engineers call parasitic 00:07:37.800 --> 00:07:40.360 capacitance. And the result is that the transistors 00:07:40.360 --> 00:07:43.500 can switch much faster and use way less power. 00:07:43.639 --> 00:07:47.199 Her whole dissertation was on this. Extreme submicrometer 00:07:47.199 --> 00:07:51.459 silicon on insulator MOSFETs. So she's deep in 00:07:51.459 --> 00:07:54.019 the physics of efficiency at the tiniest scales. 00:07:54.240 --> 00:07:56.160 And here's where we connect the physics to the 00:07:56.160 --> 00:07:58.870 strategy. She wasn't just studying some established 00:07:58.870 --> 00:08:02.050 process. She dedicated years of her life to an 00:08:02.050 --> 00:08:04.290 unproven technique. And that's a template for 00:08:04.290 --> 00:08:06.370 how she manages risk later. You know, betting 00:08:06.370 --> 00:08:08.750 the company on a new architecture like Zen, that's 00:08:08.750 --> 00:08:11.329 only possible if the CEO truly understands the 00:08:11.329 --> 00:08:13.829 science and the potential payoff of an unproven 00:08:13.829 --> 00:08:16.269 technology. SOI taught her how to manage that 00:08:16.269 --> 00:08:18.629 kind of high -stakes technical uncertainty. And 00:08:18.629 --> 00:08:20.990 that deep conviction, it fueled her professional 00:08:20.990 --> 00:08:24.089 career, starting right away in 1994. Which brings 00:08:24.089 --> 00:08:26.899 us into part two. This is the era of her big 00:08:26.899 --> 00:08:29.899 technical breakthroughs and then that quick transition 00:08:29.899 --> 00:08:32.600 into management. Her time at Texas Instruments, 00:08:32.720 --> 00:08:36.279 IBM, and Freescale. She starts at Texas Instruments 00:08:36.279 --> 00:08:39.639 in 94, right after her PhD. A member of the technical 00:08:39.639 --> 00:08:41.679 staff at their Semiconductor Process and Device 00:08:41.679 --> 00:08:44.820 Center. A classic. Deeply technical first job. 00:08:44.980 --> 00:08:48.080 But she moves fast, jumps to IBM in 1995 as a 00:08:48.080 --> 00:08:50.299 research staff member, specializing in device 00:08:50.299 --> 00:08:52.500 physics. And in just a few years, she's appointed 00:08:52.500 --> 00:08:55.440 vice president of IBM's Semiconductor R &D Center, 00:08:55.600 --> 00:08:59.019 a huge step up. Her 12 years at IBM, from 95 00:08:59.019 --> 00:09:02.440 to 2007, were really defined by solving maybe 00:09:02.440 --> 00:09:05.279 the single biggest material science challenge 00:09:05.279 --> 00:09:07.639 in semiconductors at the time. The copper problem. 00:09:07.840 --> 00:09:09.720 The copper problem. It's such a great corporate 00:09:09.720 --> 00:09:11.860 story because it's a material science paradox. 00:09:12.220 --> 00:09:14.909 Okay. Explain that. Well, for decades, the industry 00:09:14.909 --> 00:09:18.230 used aluminum for the tiny wires, the interconnects 00:09:18.230 --> 00:09:20.250 that connect all the transistors on a chip. And 00:09:20.250 --> 00:09:23.110 aluminum worked fine. It worked. But copper is 00:09:23.110 --> 00:09:25.629 so much better. It's a better conductor, which 00:09:25.629 --> 00:09:27.570 means you can run chips at much higher speeds. 00:09:28.009 --> 00:09:30.210 The whole industry was desperate to switch to 00:09:30.210 --> 00:09:33.370 copper. But there is a catch. A huge catch. The 00:09:33.370 --> 00:09:37.029 material science was a nightmare. Copper, while 00:09:37.029 --> 00:09:39.809 faster, was also incredibly reactive. It was 00:09:39.809 --> 00:09:42.029 a contamination issue, right? A fatal contamination 00:09:42.029 --> 00:09:45.549 issue. If even tiny bits of copper touched the 00:09:45.549 --> 00:09:48.830 sensitive silicon, it could completely and irreversibly 00:09:48.830 --> 00:09:50.870 destroy the chip during production. It's like 00:09:50.870 --> 00:09:52.769 trying to build something delicate with a tool 00:09:52.769 --> 00:09:55.590 that melts everything it touches. A massive paradox. 00:09:56.009 --> 00:09:57.850 And she played a critical role in developing 00:09:57.850 --> 00:10:00.490 the recipe to make it work. It involved creating 00:10:00.490 --> 00:10:03.870 new barrier materials and processes to basically 00:10:03.870 --> 00:10:06.250 quarantine the copper and stop it from spreading. 00:10:06.389 --> 00:10:09.129 And the result was revolutionary. IBM launched 00:10:09.129 --> 00:10:11.730 the copper technology in 1998. It set a whole 00:10:11.730 --> 00:10:15.169 new industry standard. And yielded chips that 00:10:15.169 --> 00:10:17.429 were up to 20 % faster than the old aluminum 00:10:17.429 --> 00:10:20.190 ones. A clear, quantifiable leap in performance, 00:10:20.409 --> 00:10:22.850 all driven by her team solving that foundational 00:10:22.850 --> 00:10:24.929 materials problem. What's really telling about 00:10:24.929 --> 00:10:26.929 her mindset here is a quote from the source material. 00:10:27.210 --> 00:10:29.350 I think I know the one. She wasn't an expert 00:10:29.350 --> 00:10:31.629 in copper to begin with. She was an SOI expert. 00:10:31.909 --> 00:10:34.889 But she said she migrated to where the problems 00:10:34.889 --> 00:10:36.940 were. That's it. That's not just being good at 00:10:36.940 --> 00:10:39.740 your job. That's strategic problem solving. It's 00:10:39.740 --> 00:10:41.519 about finding the biggest bottleneck for the 00:10:41.519 --> 00:10:44.539 entire industry and just attacking it. And here's 00:10:44.539 --> 00:10:46.220 where it gets really interesting. This is the 00:10:46.220 --> 00:10:50.440 transition from pure technical genius to executive 00:10:50.440 --> 00:10:53.659 leader. After proving herself on the copper problem, 00:10:53.879 --> 00:10:56.759 her career path shifts hard towards management. 00:10:56.779 --> 00:11:00.370 A key pivot. In 2000, she gets this critical 00:11:00.370 --> 00:11:02.629 assignment as the technical assistant to IBM 00:11:02.629 --> 00:11:05.789 CEO Lou Gerstner. Which is always a proving ground. 00:11:05.950 --> 00:11:07.909 It's a fast track to see how the business side 00:11:07.909 --> 00:11:10.169 of tech actually works from the very top. And 00:11:10.169 --> 00:11:12.370 right after that, she goes from being an indispensable 00:11:12.370 --> 00:11:15.990 engineer to an entrepreneurial manager. She becomes 00:11:15.990 --> 00:11:18.090 the head and founder of the emerging products 00:11:18.090 --> 00:11:20.610 division at IBM. She was essentially told to 00:11:20.610 --> 00:11:22.429 run an internal startup. Right, from scratch. 00:11:23.039 --> 00:11:25.620 She hired just 10 people to focus on new, high 00:11:25.620 --> 00:11:27.940 -growth areas, biochips, and more importantly, 00:11:28.200 --> 00:11:31.480 low -power semiconductors. And she was, in her 00:11:31.480 --> 00:11:34.360 words, basically director of myself at the start. 00:11:35.210 --> 00:11:39.070 She had to convince IBM to invest in these risky 00:11:39.070 --> 00:11:41.450 new technologies. And their first product worked. 00:11:41.590 --> 00:11:43.990 It improved battery life in phones and handhelds, 00:11:43.990 --> 00:11:45.750 showing there was a real market there. That work 00:11:45.750 --> 00:11:48.970 got recognized almost immediately. MIT Technology 00:11:48.970 --> 00:11:52.110 Review named her a top innovator under 35 in 00:11:52.110 --> 00:11:55.210 2001 for it. And her IBM years also included 00:11:55.210 --> 00:11:58.149 this massive, high -profile collaboration that 00:11:58.149 --> 00:12:00.870 basically shaped the future of gaming. The Cell 00:12:00.870 --> 00:12:04.649 Microprocessor. Yes. She represented IBM in that 00:12:04.649 --> 00:12:07.169 partnership with Sony and Toshiba. And the goal 00:12:07.169 --> 00:12:10.730 was just audacious. Ken Kutaragi, the father 00:12:10.730 --> 00:12:12.950 of the PlayStation, set the target. He wanted 00:12:12.950 --> 00:12:15.090 to improve game processor performance by a factor 00:12:15.090 --> 00:12:17.210 of 1 ,000. 1 ,000. It was a nearly impossible 00:12:17.210 --> 00:12:19.210 jump that meant they had to throw out conventional 00:12:19.210 --> 00:12:22.240 chip design. And her team came up with this innovative 00:12:22.240 --> 00:12:24.820 architecture, a nine processor chip. Which became 00:12:24.820 --> 00:12:27.000 the legendary cell microprocessor that powered 00:12:27.000 --> 00:12:29.100 the PlayStation 3. So think about that experience. 00:12:29.360 --> 00:12:31.580 Managing radical ambition, impossible technical 00:12:31.580 --> 00:12:34.179 constraints, high stakes corporate partnerships. 00:12:34.539 --> 00:12:37.100 All skills she would need later when she had 00:12:37.100 --> 00:12:40.120 to bet AMD's entire future on the Zen architecture. 00:12:40.559 --> 00:12:43.980 Exactly. She mastered device physics and then 00:12:43.980 --> 00:12:47.259 she mastered the politics of multi -billion dollar 00:12:47.259 --> 00:12:50.320 R &D. So after becoming a vice president at IBM, 00:12:50.620 --> 00:12:54.200 she makes a big move in 2007 to Freescale Semiconductor. 00:12:54.240 --> 00:12:56.980 And at Freescale, she really cements her identity 00:12:56.980 --> 00:12:59.620 as an executive and a turnaround expert. She 00:12:59.620 --> 00:13:03.000 joins as CTO, heading up R &D, and later becomes 00:13:03.000 --> 00:13:06.360 a senior VP and GM of their networking and multimedia 00:13:06.360 --> 00:13:09.220 group. This role pushes her even further from 00:13:09.220 --> 00:13:12.200 pure research and into P &L responsibility for 00:13:12.200 --> 00:13:14.840 a major business unit. And the move showed her 00:13:14.840 --> 00:13:17.519 skills were transferable. The impact was clear 00:13:17.519 --> 00:13:19.940 enough that E! Times credited her with helping 00:13:19.940 --> 00:13:22.399 Freescale get its house in order. She was putting 00:13:22.399 --> 00:13:24.299 the technical and structural pieces in place 00:13:24.299 --> 00:13:27.120 that led to their successful IPO in 2011. So 00:13:27.120 --> 00:13:29.259 by the time she leaves Freescale, she's navigated 00:13:29.259 --> 00:13:31.759 three huge different corporate environments. 00:13:31.879 --> 00:13:34.240 Always moving up from technical problem solver 00:13:34.240 --> 00:13:37.080 to strategic group leader, she was primed for 00:13:37.080 --> 00:13:39.059 the biggest challenge of her life. Which brings 00:13:39.059 --> 00:13:42.080 us to the moment that changed everything. her 00:13:42.080 --> 00:13:44.720 move to the struggling, debt -ridden, advanced 00:13:44.720 --> 00:13:48.200 micro -devices. Part three, the AMD turnaround 00:13:48.200 --> 00:13:52.259 strategy from 2012 to 2016. This is the core 00:13:52.259 --> 00:13:55.620 of the story. She joins AMD in January 2012 as 00:13:55.620 --> 00:13:58.559 a senior VP and general manager. Right away, 00:13:58.779 --> 00:14:01.100 it's clear she wasn't hired to maintain the status 00:14:01.100 --> 00:14:03.559 quo. No, she was there to execute a strategic 00:14:03.559 --> 00:14:06.480 change because the company was, and this is not 00:14:06.480 --> 00:14:08.799 an exaggeration, on life support. They were financially 00:14:08.799 --> 00:14:12.009 fragile. way behind technically and completely 00:14:12.009 --> 00:14:14.090 dependent on the PC market. Which was collapsing 00:14:14.090 --> 00:14:16.090 because of the rise of mobile. If they stayed 00:14:16.090 --> 00:14:18.580 tied to the PC, they were going to sink. So the 00:14:18.580 --> 00:14:20.440 sources say she immediately played a prominent 00:14:20.440 --> 00:14:23.059 role in pushing AMD to diversify. That was job 00:14:23.059 --> 00:14:25.659 one, applying that emerging products philosophy 00:14:25.659 --> 00:14:29.279 from IBM. If your core business is dying, you 00:14:29.279 --> 00:14:31.840 have to innovate in new markets. And the critical 00:14:31.840 --> 00:14:34.460 wins here were the game consoles. This was just 00:14:34.460 --> 00:14:36.899 strategic brilliance. She worked directly with 00:14:36.899 --> 00:14:39.120 Microsoft and Sony to get AMD chips into the 00:14:39.120 --> 00:14:41.460 Xbox One and the PlayStation 4. And that wasn't 00:14:41.460 --> 00:14:43.740 just about revenue. It was a high -volume contract 00:14:43.740 --> 00:14:46.419 that gave them stable cash flow. It was a lifeline. 00:14:46.759 --> 00:14:48.940 It allowed AMD to survive long enough to actually 00:14:48.940 --> 00:14:51.059 fund the development of their next generation 00:14:51.059 --> 00:14:55.139 CPUs. And that shift was stunningly fast. When 00:14:55.139 --> 00:14:58.340 she joined in 2012, non -PC sales were about 00:14:58.340 --> 00:15:01.879 10 % of AMD's revenue. And by early 2015? It 00:15:01.879 --> 00:15:04.919 was roughly 40%. She took them from total reliance 00:15:04.919 --> 00:15:07.179 on a dying market to a diversified revenue base 00:15:07.179 --> 00:15:10.080 in just three years. That success proved her 00:15:10.080 --> 00:15:12.320 strategic ability to the board, and it proved 00:15:12.320 --> 00:15:14.220 to the market that she knew where future growth 00:15:14.220 --> 00:15:16.679 was. Which led directly to her appointment as 00:15:16.679 --> 00:15:20.080 CEO in October 2014. And when she took over, 00:15:20.220 --> 00:15:22.519 her plan sounded simple, but it was incredibly 00:15:22.519 --> 00:15:25.519 high risk. First, focus on the right technology 00:15:25.519 --> 00:15:28.179 investments. Second, streamline the product line. 00:15:28.480 --> 00:15:30.960 Get rid of distractions. And third, accelerate 00:15:30.960 --> 00:15:33.399 the development of new technology. Let's stick 00:15:33.399 --> 00:15:35.840 with that first point, the right technology investments. 00:15:36.279 --> 00:15:38.600 That meant one thing, betting the entire company 00:15:38.600 --> 00:15:41.799 on a brand new microarchitecture, Zen. Which 00:15:41.799 --> 00:15:44.779 required huge R &D costs at a time when AMD could 00:15:44.779 --> 00:15:47.299 least afford it. This was not about quick fixes. 00:15:47.360 --> 00:15:50.059 It was a long, painful restructuring focused 00:15:50.059 --> 00:15:52.860 on patient R &D. Exactly what she learned at 00:15:52.860 --> 00:15:56.960 IBM. In 2015, they laid out the strategy. Focus 00:15:56.960 --> 00:15:59.639 on high -performance computing and graphics for 00:15:59.639 --> 00:16:02.100 three growth areas. Gaming, which they already 00:16:02.100 --> 00:16:04.779 had a foothold in. The data center, where they 00:16:04.779 --> 00:16:07.220 had almost no presence. And immersive platforms. 00:16:07.720 --> 00:16:10.659 The future. The data center focus was the big 00:16:10.659 --> 00:16:14.179 one. That market was exploding, and it was completely 00:16:14.179 --> 00:16:17.279 dominated by Intel. To even compete there, they 00:16:17.279 --> 00:16:19.740 needed something revolutionary. And the technical 00:16:19.740 --> 00:16:22.059 foundation for that comeback started being laid 00:16:22.059 --> 00:16:25.320 in 2016. She announced they were working on new... 00:16:25.629 --> 00:16:28.389 FinFET based chips. This was it. This was them 00:16:28.389 --> 00:16:31.970 building the core Zen micro architecture that 00:16:31.970 --> 00:16:33.889 would either save the company or bankrupt it. 00:16:33.909 --> 00:16:35.649 They are betting everything on being able to 00:16:35.649 --> 00:16:38.409 deliver a fundamentally better chip than a rival 00:16:38.409 --> 00:16:40.490 that seemed untouchable. And remember, Intel 00:16:40.490 --> 00:16:43.850 was massive, stable, deeply entrenched everywhere. 00:16:44.149 --> 00:16:46.370 To challenge them, AMD didn't just need a good 00:16:46.370 --> 00:16:48.750 chip. They needed a chip that was demonstrably 00:16:48.750 --> 00:16:51.389 faster, cheaper, and architecturally superior. 00:16:51.710 --> 00:16:53.769 And that requires incredible conviction from 00:16:53.769 --> 00:16:56.009 the CEO, knowing the payoff might not come for 00:16:56.009 --> 00:16:58.370 five or six years. But the market started to 00:16:58.370 --> 00:17:01.610 respond even before the chips arrived. AMD stocks 00:17:01.610 --> 00:17:05.049 spiked in July 2016. Fortune directly attributed 00:17:05.049 --> 00:17:08.109 that to her execution on the comeback plan. The 00:17:08.109 --> 00:17:09.849 market believed she could deliver the technical 00:17:09.849 --> 00:17:11.890 goods because of her track record. Because of 00:17:11.890 --> 00:17:13.630 the copper problem, because of the cell chip. 00:17:13.769 --> 00:17:16.940 Her credibility. built over decades in the technical 00:17:16.940 --> 00:17:19.500 trenches, was now paying off on Wall Street. 00:17:19.680 --> 00:17:22.059 It was a signal of confidence that management 00:17:22.059 --> 00:17:24.660 could actually execute this revival, not just 00:17:24.660 --> 00:17:27.180 talk about it. Which brings us finally to part 00:17:27.180 --> 00:17:30.779 four, the payoff, the rise in era, market dominance, 00:17:31.000 --> 00:17:33.160 and all the recognition that followed from 2017 00:17:33.160 --> 00:17:36.259 to today. This is where the seeds planted during 00:17:36.259 --> 00:17:39.660 the turnaround finally spectacularly bore fruit. 00:17:40.160 --> 00:17:42.420 The launch of the Zen chips, marketed as Ryzen 00:17:42.420 --> 00:17:45.079 CPUs, started in the second quarter of 2017. 00:17:45.660 --> 00:17:47.660 This was the moment of truth. And the success 00:17:47.660 --> 00:17:50.420 was immediate and incredibly disruptive. Post 00:17:50.420 --> 00:17:54.200 -launch, AMD CPU market share just surged. It 00:17:54.200 --> 00:17:56.819 hit nearly 11 % quickly and just kept climbing. 00:17:57.000 --> 00:17:58.839 But it's not just the market share number. The 00:17:58.839 --> 00:18:00.700 products themselves were genuinely competitive. 00:18:00.839 --> 00:18:04.109 In some cases, far superior. Ryzen got amazing 00:18:04.109 --> 00:18:06.809 reviews, specifically for hitting two things 00:18:06.809 --> 00:18:08.930 Intel was struggling with at the time. High thread 00:18:08.930 --> 00:18:11.630 count, so more processing capability and prices 00:18:11.630 --> 00:18:14.210 that were way lower than Intel's. For consumers 00:18:14.210 --> 00:18:17.089 building high performance PCs, AMD suddenly had 00:18:17.089 --> 00:18:19.490 a much better value proposition. And this was 00:18:19.490 --> 00:18:21.869 partly AMD's brilliance, but it was also partly 00:18:21.869 --> 00:18:24.509 Intel's stagnation. They'd been making small 00:18:24.509 --> 00:18:26.829 incremental improvements and got stuck on their 00:18:26.829 --> 00:18:29.549 manufacturing process. It left a window open 00:18:29.549 --> 00:18:32.859 and Ryzen just blew it off the hinges. And Sue 00:18:32.859 --> 00:18:35.339 didn't just stop with consumers. She immediately 00:18:35.339 --> 00:18:38.240 drove AMD back into the high -performance market 00:18:38.240 --> 00:18:40.460 with the Threadripper workstation processors. 00:18:40.859 --> 00:18:43.200 Which was a critical move. It grabbed the attention 00:18:43.200 --> 00:18:45.519 of the most demanding users, developers, engineers, 00:18:46.000 --> 00:18:48.380 data scientists who set the tone for the market. 00:18:48.920 --> 00:18:52.160 So with the core CPU business revived, her focus 00:18:52.160 --> 00:18:55.099 pivoted again toward the most lucrative future 00:18:55.099 --> 00:18:58.160 market, the data center and AI. Which led to 00:18:58.160 --> 00:19:00.859 a massive consolidation move that redefined AMD's 00:19:00.859 --> 00:19:03.640 future. The acquisition of Z -Lynx in February 00:19:03.640 --> 00:19:08.180 2022 for a reported $49 billion. This was a huge 00:19:08.180 --> 00:19:11.359 statement. AMD was no longer just a CPU and GPU 00:19:11.359 --> 00:19:14.279 maker. They were a full stack. high -performance 00:19:14.279 --> 00:19:16.160 computing company. And for you, the listener, 00:19:16.299 --> 00:19:19.339 we need to break this down because FPGAs, field 00:19:19.339 --> 00:19:22.720 programmable gate arrays, are essential for modern 00:19:22.720 --> 00:19:26.440 AI. So why spend $49 billion on a company that 00:19:26.440 --> 00:19:29.420 makes these programmable chips? Because traditional 00:19:29.420 --> 00:19:33.279 CPUs and GPUs have fixed architectures. FPGAs 00:19:33.279 --> 00:19:35.359 are different. Their internal circuits can be 00:19:35.359 --> 00:19:37.559 completely reprogrammed after they're manufactured. 00:19:37.779 --> 00:19:39.539 So you can optimize them for a very specific 00:19:39.539 --> 00:19:42.920 task. Exactly. And in AI, where algorithms change 00:19:43.049 --> 00:19:45.569 every single month, that flexibility is priceless. 00:19:45.829 --> 00:19:48.210 They offer super low latency and customizability 00:19:48.210 --> 00:19:51.269 for data center and AI workloads. So this acquisition 00:19:51.269 --> 00:19:53.990 gives AMD a critical asset for winning massive 00:19:53.990 --> 00:19:56.349 contracts in AI and other specialized fields. 00:19:56.549 --> 00:19:58.970 And it directly enables her 2025 recognition 00:19:58.970 --> 00:20:02.349 as an architect of AI. It means AMD now controls 00:20:02.349 --> 00:20:05.509 the full stack from fixed CPUs and GPUs to flexible 00:20:05.509 --> 00:20:08.730 FPGA accelerators. After that deal closed, she 00:20:08.730 --> 00:20:11.789 formally became the chair of AMD's board, consolidating 00:20:11.789 --> 00:20:14.309 her vision. for this new massive entity. And 00:20:14.309 --> 00:20:17.670 of course, that level of success brings incredible 00:20:17.670 --> 00:20:20.509 personal financial recognition. She was famously 00:20:20.509 --> 00:20:23.690 the highest paid CEO on the S &P 500 in 2019. 00:20:24.130 --> 00:20:27.890 Her pay package was $58 .5 million. And it's 00:20:27.890 --> 00:20:29.549 important to connect that to the bigger picture. 00:20:29.730 --> 00:20:32.630 That massive number was mostly a stock reward 00:20:32.630 --> 00:20:35.730 tied directly to the exponential growth of AMD's 00:20:35.730 --> 00:20:38.130 valuation. It's the market rewarding the value 00:20:38.130 --> 00:20:41.390 she created. She took a $3 billion company and 00:20:41.390 --> 00:20:44.130 turned it into a $200 billion powerhouse. The 00:20:44.130 --> 00:20:46.849 financial reward is just a mirror of that shareholder 00:20:46.849 --> 00:20:49.549 value creation. And the awards followed. Not 00:20:49.549 --> 00:20:51.970 just financial success, but deep respect from 00:20:51.970 --> 00:20:54.799 her peers in the industry and academia. There 00:20:54.799 --> 00:20:57.079 are too many to list, but a few really stand 00:20:57.079 --> 00:20:59.299 out. Like being named Executive of the Year by 00:20:59.299 --> 00:21:01.880 E! Times in 2014, right at the start of the turnaround. 00:21:02.279 --> 00:21:04.359 That shows they saw the potential immediately. 00:21:04.740 --> 00:21:07.000 But the most important one, for me, connecting 00:21:07.000 --> 00:21:10.420 back to her roots, came in 2021. She became the 00:21:10.420 --> 00:21:13.140 first woman to receive the I .E. Robert N. Noyce 00:21:13.140 --> 00:21:15.230 Medal. Which is one of the highest honors in 00:21:15.230 --> 00:21:17.710 the entire semiconductor industry. It confirms 00:21:17.710 --> 00:21:20.210 that her peers recognize her original technical 00:21:20.210 --> 00:21:23.509 contributions, the SOI and copper work, just 00:21:23.509 --> 00:21:25.569 as much as her business success. Her election 00:21:25.569 --> 00:21:27.829 to the National Academy of Engineering in 2018 00:21:27.829 --> 00:21:30.490 is another nod to that technical foundation. 00:21:30.829 --> 00:21:35.089 And now her influence is global. In 2025, Forbes 00:21:35.089 --> 00:21:37.849 named her the 10th most powerful woman in the 00:21:37.849 --> 00:21:39.890 world. We should also mention the recognition 00:21:39.890 --> 00:21:42.630 that goes beyond the balance sheet. In 2022, 00:21:42.890 --> 00:21:45.509 she was an International Peace Honors honoree. 00:21:45.529 --> 00:21:48.410 For her work donating supercomputing power for 00:21:48.410 --> 00:21:50.609 infectious disease research during global health 00:21:50.609 --> 00:21:53.569 crises. It ties technical excellence to real 00:21:53.569 --> 00:21:56.809 societal contribution. It's no wonder MIT chose 00:21:56.809 --> 00:22:00.269 to honor her. That's right. In 2022, MIT named 00:22:00.269 --> 00:22:02.789 its new nanotechnology research building, Building 00:22:02.789 --> 00:22:05.990 12, in her honor. Think about that. She starts 00:22:05.990 --> 00:22:08.250 as an undergrad making test wafers in a lab, 00:22:08.329 --> 00:22:10.630 and now the entire nanotechnology building is 00:22:10.630 --> 00:22:12.839 named after her. That is the definition of coming 00:22:12.839 --> 00:22:15.279 full circle. It permanently links her legacy 00:22:15.279 --> 00:22:17.839 to the foundational science she mastered. OK, 00:22:17.920 --> 00:22:19.759 before we wrap up, let's touch on some personal 00:22:19.759 --> 00:22:22.700 context and a fascinating little connection that 00:22:22.700 --> 00:22:24.980 adds some color to this whole story. Part five, 00:22:25.140 --> 00:22:27.900 the human context behind the corporate drama. 00:22:28.180 --> 00:22:31.180 Lisa Su is based in Austin, Texas, with her husband, 00:22:31.339 --> 00:22:34.380 Daniel Lin. And the sources mentioned an interesting 00:22:34.380 --> 00:22:37.210 habit she picked up after becoming CEO. She started 00:22:37.210 --> 00:22:39.829 boxing regularly with a trainer. Which you often 00:22:39.829 --> 00:22:42.349 hear about with highly driven executives, right? 00:22:42.490 --> 00:22:44.789 It's not just a workout. No, boxing requires 00:22:44.789 --> 00:22:48.150 intense discipline, quick reactions, mental toughness, 00:22:48.309 --> 00:22:51.170 all qualities that mirror her corporate strategy 00:22:51.170 --> 00:22:54.170 against Intel. And as of 2024, her net worth 00:22:54.170 --> 00:22:57.200 is over a billion dollars. partly thanks to the 00:22:57.200 --> 00:23:00.059 generative AI stock boom. Which just reinforces 00:23:00.059 --> 00:23:02.500 that her big strategic bets, like buying Zilinx, 00:23:02.599 --> 00:23:05.619 position AMD perfectly for the AI era. But now 00:23:05.619 --> 00:23:07.579 for the most compelling piece of trivia in the 00:23:07.579 --> 00:23:09.920 sources. The family connection. The family connection. 00:23:10.460 --> 00:23:13.619 NVIDIA co -founder and CEO Jensen Huang is her 00:23:13.619 --> 00:23:15.700 first cousin. It's an incredible detail. The 00:23:15.700 --> 00:23:17.940 source clarifies the relationship. Su's maternal 00:23:17.940 --> 00:23:20.180 grandfather is the eldest brother of Huang's 00:23:20.180 --> 00:23:23.599 mother. So let's process this. The leaders. Of 00:23:23.599 --> 00:23:26.299 two of the biggest rival semiconductor giants 00:23:26.299 --> 00:23:29.779 on the planet. AMD and Nvidia. The companies 00:23:29.779 --> 00:23:32.140 literally building the infrastructure for AI, 00:23:32.359 --> 00:23:34.480 for gaming, for the modern data center. Of our 00:23:34.480 --> 00:23:36.940 first cousins. It adds this whole other layer 00:23:36.940 --> 00:23:39.460 to the intensity of the industry. Their companies 00:23:39.460 --> 00:23:42.140 are in constant competition and that rivalry 00:23:42.140 --> 00:23:45.099 literally runs through the family tree. Imagine 00:23:45.099 --> 00:23:48.000 those holiday dinners. It just perfectly encapsulates 00:23:48.000 --> 00:23:50.759 the high stakes, fast moving world of Silicon 00:23:50.759 --> 00:23:53.359 Valley power. And finally, we should note her 00:23:53.359 --> 00:23:55.579 ongoing involvement at the industry policy level. 00:23:55.759 --> 00:23:57.839 She sits on the board of the U .S. Semiconductor 00:23:57.839 --> 00:24:00.799 Industry Association. So her trajectory is complete, 00:24:00.980 --> 00:24:04.440 from deep academic research to operational excellence, 00:24:04.680 --> 00:24:07.559 corporate revival, industry leadership, and now 00:24:07.559 --> 00:24:10.400 policy guidance. It's time for the synthesis. 00:24:10.880 --> 00:24:13.700 Lisa Su's journey, it just demonstrates that 00:24:13.700 --> 00:24:16.880 deep foundational technical expertise from SOI 00:24:16.880 --> 00:24:19.460 to copper interconnects. When you combine that 00:24:19.460 --> 00:24:22.500 with a bold strategic vision. That is the formula 00:24:22.500 --> 00:24:25.059 for corporate transformation. She went from solving 00:24:25.059 --> 00:24:27.920 intricate device physics problems measured in 00:24:27.920 --> 00:24:30.859 submicrometers to solving monumental business 00:24:30.859 --> 00:24:33.680 challenges measured in tens of billions of dollars. 00:24:33.880 --> 00:24:36.700 Her ability to diversify AMD away from the PC 00:24:36.700 --> 00:24:38.880 was only possible because she understood the 00:24:38.880 --> 00:24:41.099 technical architecture needed to win in gaming 00:24:41.099 --> 00:24:44.019 and the data center. She knew exactly what AMD's 00:24:44.019 --> 00:24:46.099 technical flaws were, and she knew what kind 00:24:46.099 --> 00:24:49.180 of massive existential bet was needed to create 00:24:49.180 --> 00:24:52.279 something completely new like... You cannot lead 00:24:52.279 --> 00:24:54.900 a true technical revival without understanding 00:24:54.900 --> 00:24:57.519 the technology at the component level. And you 00:24:57.519 --> 00:25:00.180 need the conviction to stick with that R &D plan 00:25:00.180 --> 00:25:02.880 through years of financial pain. And that conviction, 00:25:03.079 --> 00:25:05.799 born in her MIT days, is what allowed her to 00:25:05.799 --> 00:25:08.180 make those long -term bets that eventually resulted 00:25:08.180 --> 00:25:10.759 in the overwhelming success of Ryzen. She wasn't 00:25:10.759 --> 00:25:12.859 hoping a new marketing campaign would save AMD. 00:25:13.019 --> 00:25:15.799 She knew only superior physics could do it. Which 00:25:15.799 --> 00:25:18.019 brings us to our final provocative thought for 00:25:18.019 --> 00:25:20.819 you, the listener. Her early work, her esoteric 00:25:20.819 --> 00:25:23.500 PhD research, was all about making transistors 00:25:23.500 --> 00:25:26.119 more efficient using silicon -on -insulator technology, 00:25:26.599 --> 00:25:29.920 a deep focus on maximizing performance while 00:25:29.920 --> 00:25:32.880 minimizing power leakage and waste heat. So now, 00:25:32.940 --> 00:25:35.200 considering her recent recognition as an architect 00:25:35.200 --> 00:25:38.380 of AI, which is the most demanding, power -hungry 00:25:38.380 --> 00:25:41.339 technology of our time, what stands out to you 00:25:41.339 --> 00:25:43.980 about how that decades -long focus on fundamental 00:25:43.980 --> 00:25:47.480 efficiency now directly enables this global computational 00:25:47.480 --> 00:25:50.019 hunger? It shows how a 10 -year -old's curiosity 00:25:50.019 --> 00:25:53.500 about how things work can eventually drive the 00:25:53.500 --> 00:25:56.480 entire infrastructure of global computation decades 00:25:56.480 --> 00:25:59.059 later. It's the ultimate lesson in foundational 00:25:59.059 --> 00:26:01.859 research and patience. The esoteric physics work 00:26:01.859 --> 00:26:04.539 of yesterday is not just academic. It's the trillion 00:26:04.539 --> 00:26:06.619 -dollar infrastructure of tomorrow. Exactly. 00:26:07.220 --> 00:26:09.220 That's it for this deep dive. Thank you for joining 00:26:09.220 --> 00:26:11.420 us as we unpack the incredible career and strategic 00:26:11.420 --> 00:26:13.500 brilliance of Lisa Su. We'll see you next time.