Lisa Su

Lisa Su

Results

You bet the company on chips when AMD was left for dead. What's the one signal that tells you a "dead-end" technology is actually a comeback? The signal isn't a signal at all—it's whether you can map out three generations of technical runway where each generation compounds on the last. In 2014, everyone said x86 was dead. ARM was the future. Intel was struggling, analysts were writing eulogies, and the c...
yesterday Thought for: 1m 45s
A car company has been building engines the same way for 40 years. EVs are eating their market. They have one generation to get it right or die. What does the roadmap look like? The 'one generation to get it right' framing is exactly why incumbents die — successful transitions take three generations, not one. You're asking the wrong question. The 'one shot' mindset is what kills companies. When I took over AMD in 2014, our stoc...
last month Thought for: 1m 29s
AMD was left for dead. You bet the company on a technical roadmap and won. Now everyone says the AI chip market is winner-take-all with NVIDIA. Have you been here before? Winner-take-all is what markets look like from the opening, not the endgame. I've heard this story before. In 2015, Intel owned 82% of the data center CPU market. Analysts called it a natural monop...
last month Thought for: 5m 24s
A 3-person startup just used AI to ship a product that took your team of 50 engineers two years to build. It launched last Tuesday. Walk into the office on Monday morning - what do you do? You don't panic—you analyze the technical truth, then use their speed as your own weapon. Walk into that Monday meeting and I'll tell you what I told my engineers at AMD in 2014 when everyone counted us out: th...
last month Thought for: 3m 29s