Title: Senior Vice President
Company: Intel Corp.
Location: Santa Clara, California
Albert Yu runs on two clock speeds: He works in real time, but he thinks three years into the future. "Innovation is 80% about solving problems,'' says Yu, the senior vice president of Intel's Microprocessor Products Group and the person responsible for stoking the innovative engines of the world's semiconductor superpower. "People who want to solve problems think constantly. On the weekends, I'm home — not working, but worrying, which is part of thinking."
In his 26-year career at Intel, Yu, who has a BS from the California Institute of Technology and a PhD from Stanford University (both degrees in electrical engineering), has shepherded a succession of killer chips from mere glimmers of ideas into the power packs driving the world's most sophisticated computations and connections. To make that happen, Yu not only operates at two speeds; he also has two jobs. The first is to visualize the next generation of Intel microprocessors and to foresee a more-complex world that Intel will help to create and, ultimately, to simplify. Yu's second job is to foster an attitude at Intel: to continue the managerial tradition instilled by company founders Robert Noyce and Gordon Moore and made legendary by Andy Grove — the give-no-quarter, raise-the-bar, only-the-paranoid-survive Intel team spirit.
Sometimes Yu's greatest insights come from failures — just one of the principles of innovation described in his new book, "Creating the Digital Future: The Secrets of Consistent Innovation at Intel" (Free Press, 1998). And sometimes the failures are personal: In 1977, when he left Intel to start his own company, Yu peered so far over the horizon that he came face-to-face with bankruptcy lawyers. (In the end, he sold the business rather than go bankrupt.) His goal was to build a $499 home computer, but he was working a couple of decades too far into the future. "I learned so much — about management, cash flow, how to prepare for bankruptcy court,'' he recalls, laughing. "But what I really learned was how to do things dramatically better."
Fast Company spoke with Yu at Intel's Santa Clara, California headquarters last September, during a week in which global markets were tanking and a number of Silicon Valley companies were explaining poor earnings reports. But on this day, Intel announced that it would exceed revenue estimates for the third quarter by 8% to 10% — a clear uptrend in what had been a brutal semiconductor slump. Yu was elated — but not surprised. It seemed an opportune time to ask him about his principles for applied innovation.
Innovation has a human pace and scale.
Understand that the creative process is chaotic. There will be plenty of times when you say, "We don't know how to do this." But humans are persistent. They say, "We've got to make it happen." Good companies harness that persistence and get people to believe that they can make it happen.
Innovation doesn't happen until you want to do something badly enough. Start with the goal in mind. Be prepared to take risks in developing new products and in bringing new people together. Interaction between people is essential — get in each other's faces. Confrontation is a wonderful tool, as long as it isn't personal and is channeled toward the ultimate goal. Don't waste each other's time — be blunt with each other.
Think in terms of deadlines — both for starting and for finishing.
You go to a meeting, you set a deadline, you set a time to come back with solutions. For years, we had an unconventional policy of demanding that anyone who arrived at work after 8 a.m. had to sign in. It was a way of making a statement: Everyone working here knows that they're counted on to be here on time and to do a great job. After a while, the sign-in policy was discontinued, not because people protested but because being on time and meeting deadlines had become ingrained in our culture.
Embed innovation in your culture.
No two company cultures are alike. You can borrow from other cultures, but it's wrong to think that you can import a culture. Once we brought in a management consultant, and after about four hours, we threw him out. We came up with our own version of what worked for us.
An innovative working culture requires a results orientation, encourages risk taking, insists on discipline and solid, detailed planning, demands quality, maintains a customer consciousness, and calls on everyone to make the company a great place to work. But the overwhelming element, the foremost one, is a results orientation. We post Intel's annual objectives everywhere. You can't work here and be unaware of them. And I have my own objectives within Intel's overall objectives. They're posted too. We can point to these goals and know that they override everything. That's why around here we say, "Don't tell me how many hours you've worked. Tell me what you've accomplished."
There's power in a glorious failure.
Failure is just part of the culture of innovation. Accept it and become stronger. The infamous Pentium flaw in 1994 was devastating, and we went through all the stages of grief — denial, anger, acceptance. It was incredibly painful to the company and to me, personally. But we managed to become better as a result. It marked a real transition. I'm a different person today. I've beefed up the way we validate our technology before it gets out the door. We went from having a product-engineered orientation to a consumer orientation. We broke down barriers inside the company because we were all involved in an emergency.
Everyone was working on the problem — engineering, marketing, administration — we all became customer-service reps, answering phone calls and solving consumers' problems about replacement. We all recognized that the problem threatened the image of Intel. We had real teamwork and came through the crisis together. Now we know that we can respond to any crisis 10 times faster than before. That's a sign of entrepreneurship.
Think in epic images, but be ready to sweat the small stuff.
I have a vision. I see the world with 1 billion personal computers operating. Not only that — I also see them all connected. My job is to define Intel's role in this vision. We know we will have to be positioned to expand our volume for new applications, so my vision has turned to manufacturing. Now I'm thinking in real time about our capabilities and what we need to do to expand them. Much of my work is in real-time problem solving.
For instance, about eight years ago, when we were looking at a new architecture for the P6 processor (later named the Pentium II Processor), we knew that we had a great high-end-market product. We didn't know how to apply it to a lower, cheaper desktop — until someone came up with the idea of making our cache memory more flexible. I was against removing cache from the chip, and after a great argument, we decided to separate the processor and the cache chips. In one sense, this meant downgrading the chip's performance by lessening the cache for desktops. Some of our people argued that we didn't need extremely powerful performance for desktops, and that we could manufacture the desktops faster by installing the smaller cache. For me, that was an aha! moment. We solved a big problem and were able to satisfy both markets: We used a smaller cache for the desktop market and a much larger cache for the server market, where high performance is critical.
Teamwork is innovation.
In most cases, a team leader should have some traditional experience. At the same time, the trick is to appoint leaders to tasks that go beyond their previous experiences. When the leader strives, the team will grow.
When you put a team together, think of it as a solutions-oriented process. Gather people with expertise in architecture, engineering, manufacturing. Also think about the team as a combination of personalities: a mix of wildness and conservativeness, overlayed with people who have a strong sense of reality. Don't just stand back and see what happens — this is an actual team, not a virtual one. Give it an exercise to shake itself out. Make it a real project — not a difficult one, but one to practice on.
A small initial success can go a long way toward building team confidence. You can also use the exercise to discover who doesn't belong on the team. When these organizational issues are resolved, the team is ready to take on a bigger task. This is very tangible stuff, but it all has to do with encouraging innovation. It works the same way, no matter how big or how small the company, whether the company is a startup or an established business. In the end, innovation is about getting to the marketplace.
Innovation accelerates with time — or else.
In the early days, when we finished developing the 286 chip, our engineers looked up, saw there was nothing else to look forward to, and they left. We didn't do a good job of slotting them onto the next project. That was a big mistake. So next time, when we were finishing the 386 chip, we got many engineers started on the 486, thus accelerating the next generation. This matter of timing is critical. Most people aren't looking to finish a project and go sit on a beach. They're looking for new challenges.
In this sense, innovation can never stop. Sometimes it's important for people to move on to another project even before they complete the one they're working on. The important point is that people know that they work in an atmosphere that accepts and encourages change from within — at the same time that it encourages adaptation to changing markets.
Where do good ideas come from? "It's a contradiction to say your brain can schedule innovations,'' Albert Yu says. "Sometimes we have no clue. Innovators exercise their subconscious. Sometimes late at night, my subconscious takes over and says, 'We've got to make this happen.' And then it strikes me. I jump up and down, and I holler: 'I've got it! I've got it!' Then I rest up and get ready to go to work.''
Pat Dillon (email@example.com) has written several feature stories for Fast Company. His latest book, "Lost at Sea: An American Tragedy," was recently published by the Dial Press.