Run Toward Problems
By 2014, one of America's oldest chip companies had so little of the data center market left that executives rounded its share down to zero. Zero. The company was AMD.
AMD had already cut a quarter of its workforce, sold its Austin campus and leased it back, and spun out the chip factories it was built on. In 2012 it lost more than $1 billion in a single year, and an analyst publicly called it uninvestable. It would lose money every year from 2012 through 2016.
Revenue fell 28% in 2015, to $3.99 billion, on a net loss of $660 million. That July the stock touched $1.61 (yes, $1.61). The company carried about $2 billion in debt. At the end of 2014, all of AMD was worth $2.07 billion on the public market.
Lisa Su became CEO on October 8, 2014, and called it her dream job. She had three electrical engineering degrees from MIT and 13 years at IBM behind her, plus advice from a mentor there she never dropped: "You should run toward problems."
Her plan came down to three things: make great products, rebuild trust with customers, and simplify the business. Some board members wanted AMD to switch to low-power mobile processors, since PCs were shrinking and smartphones were taking off. Su said no. Her line to Time years later was very simple: "We needed to bet on what we were good at."
Nvidia, the company AMD is now chasing in AI, is run by her first cousin once removed. Su's maternal grandfather is the eldest brother of Jensen Huang's mother. They did not grow up together and did not meet until an industry event, well into their careers. Her line to Bloomberg: "No family dinner. It's an interesting coincidence."
Today AMD is worth more than $800 billion. Intel is worth about $470 billion. She inherited a broken balance sheet and one thing the company was good at. She decided in her first year what all of it was for, and then she stopped re-deciding.
