Editor’s Note — June 10, 2008 —
references to the IEEE Emanuel R. Piore awardees and Institute of Electronics and Electrical Engineers (IEEE) Emanuel R. Piore Foundation award fund were updated post-publication.
Rick Rashid, Senior Vice President, Research
ANAHEIM, Calif. – June 10, 2008 – This week Rick Rashid, founder and senior vice president of Microsoft Research, joins the ranks of IEEE Emanuel R. Piore awardees, having received the honor at the 45th Design Automation Conference in Anaheim.
Sponsored by the the Institute of Electronics and Electrical Engineers (IEEE) Emanuel R. Piore Foundation award fund, the Piore Award is designed to recognize leadership and accomplishments in the field of technology. Since 1977, the award has been given periodically to industry figures who have made a lasting impact with their work.
Rashid’s specific citation is for his work in the design of modern operating systems, as well as “innovation leadership in industrial research.” According to Microsoft Chairman Bill Gates, however, the award speaks not so much to what Rashid has accomplished, but to what’s possible with a career in computer science.
“I have always admired Rick’s dedication to the idea that pushing the boundaries of computer science will lead to advances that change people’s lives for the better,” says Gates. “When we hired Rick to launch Microsoft Research 17 years ago, I had high hopes for what he would achieve. Since then, his contributions to Microsoft, to the field of computer science, and to society in general have always exceeded my expectations. This well-deserved award is a fantastic acknowledgement of the important impact of his great work.”
In accepting the accolade, Rashid also nods to the influence each individual can have in the course of their career, as well as the potential to make a difference with technology.
“We lose track of the fact that individuals can make a huge impact in any field,” he says. “Awards that recognize individuals are a way of showing how a single person can help change an industry with their work, with their energy, and with their intellect.”
A Founding Influence
Rashid has invented technologies and published papers in fields ranging from networking, programming languages, distributed systems and operating systems, to computer vision, databases and cryptography.
But getting in at the ground floor and building a team has also been a hallmark of Rashid’s career since graduate school, when he took an unusual risk, eschewing a grad program in math at Berkeley to help found a new computer science department across the country at the University of Rochester.
“One of the professors, Jerry Feldman, was going to Rochester to start the department,” Rashid says. “He contacted me and told me they would have access to Xerox Alto and would be the only place in academia to have that.”
Indeed, Feldman’s group was the first outside of Xerox with access to the original Xerox Alto and the then-experimental Ethernet. According to Feldman, Xerox executives in Palo Alto wanted to get some of their machines running in proximity to the company’s headquarters in Rochester as a way to convince senior executives of the new computer’s viability. Feldman was able to use the commitment from Xerox to help recruit a new department of faculty and grad students to Rochester.
“I contacted Rick about this idea of starting a grad school at this new place that didn’t exist,” says Feldman, now a professor of cognitive science at the University of California, Berkeley. “There’s a special type of personality that would view this as a good thing. So Rick was not only one of the first graduate students at Rochester, but was a key part of the team that brought in these revolutionary computers and got software on them.”
Part of the team’s success at Rochester, according to Rashid, was Feldman’s philosophy of giving students free reign to explore and do the kinds of research they were interested in.
“I had an opportunity there to do anything,” says Rashid. “Nothing was standing in our way. We built hardware, we built operating systems, we worked on early computers, and it gave me a leg up in terms of being able to do things that a lot of people didn’t have an opportunity to do.”
During that time Rashid also developed one of the first computer games, Alto Trek, which became a standard-setter for games development using an Ethernet like network.
“We developed as part of that some new techniques for effectively doing multicast on the old three megabit Ethernet, and those ideas led in part to the multicast standards for the 10 megabit Ethernet which has then been preserved in later versions,” he says.
Rashid’s PhD thesis at Rochester was also a milestone of sorts, a very early work on the computational interpretation of moving images called “moving light displays,” which has been referenced and built upon by researchers in both computer science and human cognition.
“To be honest my work in computer vision was mostly what I did for fun when I wasn’t working on operating systems,” says Rashid.
After his time at Rochester, Rashid became a professor at Carnegie Mellon University, spending a decade at the school researching operating systems. In a familiar theme of his career, Rashid would start the University’s work in networking and distributed systems.
“When I first arrived there they were just installing their Ethernet, and they were just getting their first Xerox Altos,” he says. “So again I was in there on the ground floor. I got to train my own grad students and staff, and really build things from the ground up. So many times I’ve been given an opportunity to do things that in some sense were like a dream come true. These things drop in your lap, and hopefully you’re able to run with them.”
Rashid led a team at Carnegie Mellon University in the late 1970s and early 1980s that developed “Mach” — An operating system “microkernel” that could be used in both distributed computing environments and large-scale multiprocessors.
Mach proved to be an unusually long-lived technology. Today it is a grandparent of many commercial operating systems currently in use, including many versions of UNIX, the Mac OS and even Windows, and is still the standard by which such technologies are measured.
“Part of the culture at Carnegie Mellon was to be users for each others’ systems,” says Jeannette Wing, a CMU computer science professor who is currently serving as the assistant director for the Computer and Information Science and Engineering Directorate at the National Science Foundation. “We all used Mach and helped Rick’s team debug their experimental but robust operating system. Soon another team was building a distributed transaction facility called Camelot on top of Mach, and a colleague and I were building programming-language support called Avalon on top of Camelot. It was a perfect layering of research projects.”
During this period, Rashid not only made significant contributions to modern operating systems, but also developed a leadership and mentorship style that has served him in his years as head of Microsoft Research.
“Rick always took his entire research group to see the latest Star Trek movie, as soon as it came out,” says Wing. “He was generous with his attention to his group, and I’m sure it inspired them to work well as a team. It is this spirit of looking out for people that Rick carried to Microsoft, because I see it ingrained in the culture of Microsoft Research.”
Building Microsoft Research
In 1991, with nearly 20 years of groundbreaking research and two graduate academic startups to his credit, Microsoft Chairman Bill Gates and Nathan Myhrvold recruited Rashid to found Microsoft Research. Nearly 17 years later, Rashid is still the head of the organization he founded.
Asked why his group has thrived for nearly two decades while other corporate research organizations come and go, Rashid cites both the tight integration between Microsoft Research and the company’s broader product development mission, as well as a consistent philosophy of openness and exploration he learned from Feldman back at Rochester.
“I think we really created a new kind of research lab in industry, one that is a blend of computer science programs at universities and the industrial research lab setting,” Rashid says. “Our most important mission is moving the state-of-the-art forward in computer science, and over the years the research organization here has been incredibly well supported in that mission by the product organizations, and by Bill and Steve and the other executives. Plus, the output of research are exactly the inputs you’d imagine for Microsoft product teams, so it’s just a perfect match.”
On the other side is the well-known openness of Microsoft Research, in letting the scientists it hires pursue the technologies that interest them most.
“We don’t attempt to change or steer our researchers in any particular direction,” he says. “We try to hire the best people, give them an opportunity to be successful, in an environment where they can be successful. The most important control I have in the organization is who we hire. It’s the quality of the people that make this work.”
Rashid also says the consistency of having the same philosophy in hiring, promotion and the research agenda has been a key part of the group’s success.
“In a research environment you want people to take risks and look far out to the future, and those people need to know that the rules aren’t going to change underneath them,” he says. “They need to know that there’s going to be consistency, and I think we’ve been able to provide that.”
In the early days of Microsoft Research, Rashid says, perhaps the biggest risk his hires were taking was the decision to come to Microsoft in the first place.
“Software companies in those days weren’t seen by people in the academic community as stable places where people could get their work done,” he says. “Our researchers were pretty much taking it on faith that it would all work out. They were making a bet on the future.”
For many of those who did come, the bet has paid off. Some of the hires Rashid made in the early to mid 1990s are still at Microsoft research, having made many contributions to technology in their own rights. Names like David Heckerman, Eric Horvitz and Jim Gray became well known in the tech industry largely through contributions made during their time at Microsoft.
“Heckerman and Horvitz really founded the work we’ve done over the years in Bayesian reasoning, inference and machine intelligence and learning,” Rashid says. “Jim Gray helped to set a standard for what the lab was, and what it’s capable of. We enticed them into Microsoft as a place where they could really take their ideas and carry them much further and carry them to a broader community, and they have.”
Ling Brings Rashid Full Circle
Another key hire was Dan Ling, whom Rashid knew from his undergraduate days at Stanford. Ling went on to become a corporate vice president at Microsoft and head of the group’s Redmond lab before moving into semi-retirement a year ago.
“Rick recruited me shortly after he arrived at Microsoft,” says Ling. “When I finally visited I was incredibly impressed by the vision for Microsoft Research as well as Rick and Nathan’s passion for the lab. I believe that passion and vision have really made Microsoft Research the success that it is today.”
Interestingly enough, it was Ling who’d originally “recruited” Rashid to take his first computer science course at Stanford. Rashid took that first course, and then another architecture course working on an early mini computer called the HP 2116. At that point, he was hooked.
“This was programming the front panel switches, paper tape, teletype and that kind of thing, and I just had such a blast,” says Rashid. “The fact that my intellect could animate this piece of hardware, this physical device, just made me so excited. It was a rush when I got my first programs running on that machine. And that really hooked me, and then I went on to take every computer science course that I could at Stanford.”
It was that sense of excitement and creativity that helped Rashid take his own career risk and move across the country to help found the department at Rochester, even though careers in computer science were not mainstream at the time.
“In the early 1970s, no one knew — computer science and underwater basket weaving were probably the same general area of profession,” he says. “But if I look back on my career, I just feel like I’ve had so many opportunities in computer science. I have been privileged to do really cool things.”
In the end, according to Ling and others, from helping his professor found one of the country’s first computer science graduate schools, to helping create one of the world’s first operating systems, to creating an environment at Microsoft where other researchers could pursue their own accomplishments, Rashid’s career shows that computer science is about more than technology itself – it’s also the ability to make a difference in people’s lives.
“At Carnegie Mellon during his work on Mach, he mentored students who have gone on to very significant positions in industry and academia,” says Ling. “At a time when many thought that the best days of industry research were past, he built and led a new premier research institution in the private sector. He has worked to attract students into computer science and improve the diversity of the discipline. His message has often also focused on the tremendous opportunities to contribute to the other sciences and important social problems via technology. This human element is a big reason Rick deserves the Piore Award.”