Summer 1997
Just another computer cowboy from Nevada
Profile: Bo Ewald
|
The new executive VP of SGIWe heard the news as soon as we got to San Jose: there'd been a reorg the week before at SGI/CRI. Robert "Bo" Ewald, former president and chief operating officer of Cray Research, Inc., had been promoted to executive vice president of the Silicon Graphics Computer Systems organization. He was now responsible for desktop workstations, graphics, servers, supercomputers — all the company's core computer products. He'd be moving soon from Minnesota to California. Good-bye, lutefisk. Hello, sushi. What we wanted to know was, in the trepidant chaos that was the evolving SGI/CRI merger — who was taking over whom? By the end of the week, it seemed as if a number of Cray traditions were infusing SGI. And it looked like Bo was part of the blend. People note that Bo is nice. Politically astute, but nice. One might say that as a leader, he knows how to serve — and that his aspiration to please the customer may represent a competitive advantage for SGI in years to come.
"I've been associated with CUG since it started, really," he says. "It's been a tremendous association for me, and not just because of the computing problems to solve: the personal relations have been tremendous. I've made a lot of good friends over the last 20 years. I won't be staying away from this business — I expect to continue to be invited, if friends allow me to come." * * *
To hear tell, his career has had a somewhat logical linearity from the days of yore to the here and now — except for the cowboy stuff. It all started in Nevada
Bo Ewald was born on Nov. 20, 1947, the oldest of four children. (He has a brother in San Diego and two sisters, one in Nebraska and one in Nevada.) His mother came from an old-time Nevada family, so he grew up riding horses on a ranch in northeastern Nevada near Wells. Later, his father was in the plumbing and heating business in Reno. He attended the University of Nevada in Reno and majored in civil engineering. "The university had an IBM 1130 they used for administrative work," he recalls. "It was in the engineering building. So at night, a bunch of us juniors and a young professor interested in computing would go in and teach ourselves to program. We'd run little codes on the machine." He went on to graduate school at the University of Colorado in Boulder in 1969-70, earning a master's degree in civil engineering, then spent two years in the army. Because of his education and "a little luck," he was stationed at West Point Military Academy, where he taught computing and learned computer graphics. "It was an unbelievable assignment," he recollects, "particularly for being in the army." After West Point, Bo returned to Boulder in 1972. He was on staff at the CU computing center for the next three years, where he was involved in computer graphics research and development, applications programming, and the instructional use of computers. He also pursued a Ph.D. in civil engineering, doing all but finish the dissertation. By then he had a computer science minor — in fact, he carried more course work in computer science than in engineering. Taking to the sky
"It's a Boeing Stearman," he says. "When I visit Boeing, they say I'm the only individual they know who owns one of their planes." "A cowboy can't survive in Pennsylvania"
After his stint in Boulder at the university computing center, Bo went back to the ranch in Nevada for a year to "cowboy." "Ranch work is great," he says. "You get to be outside, there's fun stuff to do: it's good, hard, outdoor work. You name it. Fixing fences, rounding up cows, plowing fields, putting up hay, being a cowboy — there's that romantic element as well." At that point, he was ready to do something different; he wanted to go back into the real world. He'd always been in R&D, but had never spent much time in industry — so he decided to go to work for Sperry Univac in Bluebell, Pennsylvania. The job sounded interesting: managing engineering applications and developing what today are called graphics workstations. But Jim George from Los Alamos National Laboratory told him, "You'll hate it there. A cowboy can't survive in Pennsylvania. Give us a call when you're ready to come back and work at Los Alamos." "He was right," Bo laughs. "I lasted six months. Jim said, 'Great!' and offered me a job working in the computer graphics group. He said, 'Be here Monday.' I said, 'OK!' My first job was working on computer graphics research — creating visualization technology so people could see the results of their simulations." Then Los Alamos installed the first Cray supercomputer. It had no operating system. Bo had worked on high-performance graphics, and had become familiar with the Cray, so he was asked to run the Operating System Group for Cray products. Soon afterwards, the director of the Computing and Communications Division left the organization, and LANL asked Bo to take over. They were taking a chance: Bo was only 29 or 30. "It was scary," he says, "but also really fun" — and he worked with a great bunch of people. On to Cray and SGI
After four or five years, however, he was ready to try something else: "Again, I hadn't spent much time in private industry, and I'd never done sales and marketing at all. I thought it might be interesting to learn something about it. I called John Rollwagon at Cray Research and asked if there was anything for me to do. I looked at a few other jobs, but in the end I went to work for Cray in 1984. "My first job was in downtown Minneapolis when Cray headquarters were there. At Los Alamos, I'd been responsible for computing and communications, all kinds of stuff — we had a $60 million budget and 500 people. I went from running that to being a one-person band, trying to see if Cray could sell more supercomputers to universities. "I lucked out on the timing — that was when the National Science Foundation was starting the supercomputing centers program. I was called director of University Programs, something like that. We linked up with the universities to create funding for R&D. That worked out, in that universities have become the largest single segment of Cray business." Eventually, Cray Research decided to look into the possibility of selling more computers to industry, since 70 to 80 percent of the business at the time was with the government. Bo was named vice president of Commercial Marketing. ("I helped find ways to get Eli Lilly to buy computers, things like that," Bo says with a nod to Sam and Scott, both site reps from that corporation.) When consolidation was required in Cray software efforts, Bo was asked to go to Mendota Heights, a suburb of Minneapolis, to run the Software Division, which he did for several years. "Then I moved to Chippewa Falls [Wisconsin] and worked with Les Davis, where I was responsible for hardware and software research and development — although with Les, it's hard to tell who's responsible for what! Somewhere in there, I don't know when, I became president and COO." We smile. We certainly can relate to that — God knows, it's hard to keep track of small details. Bo served for about three years as president and chief operating officer of Cray Research, then the merger with SGI occurred. Subsequent to the first-year shakedown, Bo was promoted to executive vice president of SGI Computer Systems. "I started work in Mountain View last week," Bo says. "It was announced internally a week ago Friday; I started last Monday. I'll commute for a while, then get a place to rent here and move out." I ask him if he will miss Minnesota. "Ya, you bet," Bo says like a Minnesotan. He looks pensive for a moment. "I'll miss the style of living — I have a 200-acre farm, a couple of horses ..." Not to mention the landing strip. Vector processing and grand challenges
At the Cray reception the night before, I'd asked CUG attendees what questions they would like Bo to address. "Ask him about SGI/Cray's commitment to high-performance vector processing in Scalable Node architecture," said a systems analyst across the table. "Ask him what's the biggest challenge in his new job," suggested a team leader from Los Alamos. We toasted to vector processing, and we toasted to new challenges on the job. Now, as the waiter refills our coffee, I ask Bo if there will be a commitment to vector processing in the SN architecture. "Yes." Bo says, "As we look at applications, we see there are a whole lot of applications that really are best optimized around vectors. Another class of applications are best optimized around scalable systems. We will serve both communities." "What's the biggest challenge in your new job?" "My first challenge goes beyond even the Silicon Graphics/Cray arena. The challenge in supercomputing is, how do you both grow a company and be profitable? That's number one. "Second is finding out how to take the best of the two companies we have and create something even better — not take two good companies and average them. It's a cultural challenge, if you will. There are a whole set of opportunities in marketing. If you want to be number one, you have to provide a means of leadership; you must define what leading the way means. Providing a product that creates value for customers is the challenge. Product leadership is still key. "Third would be building the relationship with customers. Tied into that is the whole service process, which implies predictable software releases, escalation processes, and so forth. The bottom line is that we want to be the trusted partner of the market we serve. "Relations between the SGI/Cray teams are frankly pretty good," he adds. "The situation has improved over the past year as people get to know each other and each side earns the other's respect. The goal is to take the best of both worlds." Future vision
Flash forward — way forward. I ask Bo if he watches the science fiction TV series Babylon 5. No, he says. I ask him if he's ever heard about the Singularity — a term denoting a rapid escalation of technology and awareness toward a sudden and nontrivial phase shift in human development. He's heard of it, but not by that name. "Do you think supercomputers will save the world?" "That supposes the world is doomed," he says, and laughs. "I don't think the world is doomed. Will supercomputers create even more significant contributions to life? Absolutely. We've just scratched the surface of material science, with new compounds to make cars lighter and our lives better, with new drugs — there will be tremendous contributions in medicine. "Can supercomputers help find more petroleum, make production processes more effective? Absolutely. And it's clear that supercomputers shortened the Cold War, there's no question. "In the weather and climate research domain, it will be possible to make more accurate predictions of the cresting of the Red River; we'll be able predict conditions in real time or ahead of time in the tornado alleys. When you do that, you're trying to model chaotic systems almost at the molecular level. "Another thing I think we're going to do that we haven't done yet with the linking of SGI/CRI is take the best supercomputing and the best visualization technology between us, the best data-intensive technology — and I predict that 20 to 30 years in the future, we'll be the first company to produce a system as usable as the one on the Star Trek Enterprise — including a holodeck, the latest visualization graphics, the ability to access all the data in the universe ... It will be easy to use, we'll be able to spin through it and do scientific analyses back and forth ... we're already moving down that path." Accessing intuitionThe waiter leaves the check. Bo pays. "So ... with all your responsibilities, what do you do in your free time?" "There's a concept!" I ask if he does anything right-brained — play music? write poetry? He admits to being an amateur nature photographer.
"In your line of work, you need technical knowledge and skill to make the right decisions — but does intuition play a part?" "I draw heavily on intuition," Bo replies. "The technical basis provides the foundation, and the analytical thinking we were trained in helps; but in the end, you have to believe in what you're doing. You have to believe that what you decide is right. Sometimes you feel that much more in the pit of your stomach than in the mind. So I continually access and draw from intuition." Suddenly, time is up. We drink up, pack up, and get out. Bo Ewald, intuitive corporate computer cowboy from Nevada, must attend to other responsibilities — among them, perhaps, preparing his address for the afternoon's plenary. I remember, then, what Mary Amiot, a Cray Research liaison to CUG, told me at the reception the night before. Bo had been wondering recently if he should still come to CUG. "You'd better," she told him. "It wouldn't be CUG without you."
|
Comments to: lester@ucar.edu
