As a leader of Google’s data center development for 17 years, Joseph Kava played a pivotal role in shaping what we now know as hyperscale data centers. He and his colleagues expanded Google’s data center operations globally to gigawatt scale, while setting new standards for energy efficiency and often sharing their strategies to help the broader industry lower energy use.
Kava was recognized this week with the Data Center World Lifetime Achievement Award for his leadership at Google, where he led the company’s global data center build out to support the cloud and AI revolution. Kava received the award on April 21 at Data Center World in Washington, D.C.
Long before he could achieve any of that, however, Kava had to take a mid-career leap of faith into an area he knew absolutely nothing about.
Kava was a successful semiconductor executive at Applied Materials, 16 years into a career centered around engineering, operations, product development, and material science. But he wanted to shift out of the semiconductor industry. The company’s CFO, a mentor of Kava’s, suggested he work for the company’s CIO, learn about information technology, and then he could work in any industry.
“It never crossed my mind to work in corporate IT. And while I knew what a data center was of course, I really had no understanding of what it took to build a data center and certainly no understanding of the business side of building and operating data centers,” he says. “But I took a leap of faith and worked in the IT department and for two years learned everything I could about the whole IT stack, including data centers.”
Kava helped Applied Materials build out its data center, eventually took a COO role with a colocation provider, and was then recruited by Google, where he worked from 2008 to 2025.
Kava reflected on career lessons and key moments in a recent conversation with Chris Murphy, Content Director for Data Center World. Here are select excerpts about Kava’s leadership principles, sustainability approaches, and advice for the next generation of data center leaders.
Kava offered one leadership principle that has stayed constant throughout his career and that served him particularly well during fast-changing, hard-to-predict environments such as the AI boom:
“I've always been predisposed to action. I've told every team that I've ever led that one of my leadership principles is ‘Do something.’ Paralysis by analysis is never the right answer. Even if you're not 100% certain, do something. Be smart enough to realize that if it was the wrong thing, you can always pivot and do something different. But doing nothing is never the right answer.”
But Kava believes his leadership style has evolved:
“I think in the early days of my career leading teams, I was maybe a little bit more prone to get anxious or worried about things. When you've gone through this much growth and, frankly, leading organizations that are 24X7 forever, and leading global teams, there's always something going wrong somewhere in the world. So, you kind of develop more of a calm style. That doesn't mean that I don't take it seriously. It doesn't mean that I'm not concerned that this could be a serious thing that we have to fix right now. But I tend not to get so worked up.”
Kava reflected on how, when he joined Google, the company had just publicly shared its PUE (Power Usage Effectiveness), the key measurement of data center energy efficiency. Google was greeted with skepticism when it announced a PUE of 1.21. So, Google held a Data Center Efficiency Summit in 2009, inviting peers including AWS, Microsoft, Facebook, and other industry stakeholders, along with the media.
“We kind of pulled back the curtain and showed them our servers for the first time, showed them how they were built. We showed them our architecture of the data center in terms of our power delivery and the cooling, and we showed them all of our calculations and all the data, and we invited them to peer review everything. And it was a real aha moment, that really showed the art of what's possible. And the industry largely took a cue from that. Many of the things that we experimented with and proved out in mass production have become mainstays of design features in data centers and in server construction and technology. And that's really very gratifying.”
Asked what advice he’d give emerging leaders as they face the new challenges of the AI factory era, Kava urged people to keep questioning the status quo:
“I guess it’s to always question why we do things the way we do, because throughout the time that I've been leading data center teams, our ideas of what the right answer is have changed dramatically. And I expect that they will continue to evolve and change. So when someone says, “oh, we can't do that because of, you know, X, Y, and Z,” challenge those assumptions, because all of that is based on a certain set of constraints. And if you can find a way of opening up those constraints or removing some of them, you might find a much better, faster, cheaper way of designing, building, and operating data centers. So continue to challenge the conventional wisdom that ‘this is how we have to do it.’”