Alumni profile: A conversation with Rich Uhlig
Alum Rich Uhlig graduated from the University of Michigan with a BSE in computer engineering in 1988 and a PhD in CSE in 1995. After graduation, he completed postdoctoral fellowships in European national research labs in Germany, Greece, and France. In 1996, Uhlig joined Intel, where he is now an Intel Senior Fellow and Corporate VP.
Upon joining Intel, Uhlig has been a champion for virtualization technology, and has led the definition of multiple generations of virtualization architecture for Intel processors and platforms, known collectively as Intel Virtualization Technology (Intel VT).
Rich recently returned to campus and we spoke with him about his early interest in tech, his time at U-M, and his career at Intel.
Getting Started in Tech
“I was into computers young and got a lot of support from my folks to do that,” says Uhlig. A native of Toledo, he was exposed to computers early through programs at the University of Toledo that introduced high school and younger students to programming. He took a course there that used punch cards and a mainframe computer. “And, of course, in that era,” said Uhlig, “the PCs were just starting to come out, the TRS 80s and the Apple IIs, they were just available back then, so I just started getting into it.”
While he was in high school, Uhlig was “one of those computer geeks” who began learning programming early. He learned a variety of programming languages and progressed to understanding how the system worked at lower levels with assembly language and the underlying hardware design. And he taught courses. He turned his efforts into a sort of business almost right away, teaching classes and writing for computer magazines before graduating high school.
Experiences at Michigan
As he finished high school, there was no doubt in Uhlig’s mind about his direction – it would be in computer engineering. He was interested in a top program, so Michigan – just a short distance north – was a logical choice.
A number of formative experiences took place for him at Michigan, and when asked about which stood out his immediate response was “Meeting my wife! We met in Bursley Hall.”
He particularly valued his education in EECS, as it expanded on his previous knowledge and enabled him to work on a number of cool projects. During his graduate work, Uhlig worked on the architecture side of a project to create a gallium arsenide MIPS-compatible processor.
There was a period as Uhlig finished his CE degree where he became intensely interested in physics. “I was going to switch direction and go into grad school for physics,” said Uhlig, “but I wasn’t good enough and didn’t get in!”
He took a job at a company in Ann Arbor called Applied Dynamics International (ADI) and was able to gain a solid foundation in computer engineering. The company had begun business in the 1950s using analog computers to solve dynamic system equations, but then transitioned to a digital architecture that outpaced their competition. When Uhlig got there, he became a part of that transition and in creating what he recalls as a “really cool machine.”
“Although I had very little experience,” said Uhlig, “they put me in an architect role and I learned a lot from that and from my mentor there, Peter Bird. That really got me into computer architecture more deeply.”
Professor Trevor Mudge, an expert in computer architecture, was consulting with ADI and Uhlig connected with him there. Before long, Uhlig had applied for graduate study in CSE, with Mudge as his advisor.
According to Uhlig, “What was really great for me in my overall time at Ann Arbor was expanding my knowledge beyond CPU microarchitecture and getting into operating systems and simulation technology, just getting a more rounded view of systems, across a number domains – that was formative. Also forming a really strong research partnership with Dave Nagle; he’s a lifelong friend now as well. And Trevor Mudge. It was a foundational learning period and also a time when I forged some of those important lifetime relationships. I just love Ann Arbor and still come back here a lot.”
Overseas Postdoc Appointments
After completing his PhD, Uhlig moved to Europe and did postdocs at national research labs in Germany, Greece, and France. That was a valuable experience for him because he got exposure to more people, ideas, and viewpoints, and it was a really good thing for building a network and connections. “My heritage is Swiss and German, so I always wanted to live in Europe and improve my German, so that was a lot of fun,” said Uhlig.
“I always knew that I wanted to do research, but I also knew that I didn’t want to be a professor, so I interviewed at all the research labs in the US at the time, including at Intel.”
A Mission at Intel
Intel had begun a brand new research lab at that time, recognizing that to keep microprocessors moving forward, they had to invest from a research point of view. Uhlig was able to get in on the ground floor at the formation of that lab, and it eventually evolved and combined with other research labs at Intel.
This research lab was the environment where Uhlig was able to get started on virtualization. His first project was to do a platform simulator, a high-speed one that was able to boot operating systems at near real-time. That project, combined with his interest in operating systems and how to support them with architecture, is what got him into the virtualization space.
“From a career accomplishment point of view,” said Uhlig,” the virtualization technology is something that I am really proud of. There were others in the space at that time, and we all saw the same opportunity. Some of them got to market a little earlier because they were doing software-only solutions, but in the end, all of the stuff converged on the hardware support that Intel provided and that I started inside the company and so now everything that is virtualization does run on the architecture that I led and brought to the industry. Just seeing all of the different ways it is used is really gratifying.
“Now, I also have a great job in that I can take the things I learned and lead this larger research organization of Intel Labs in a very practical way. By practical, I mean that I think we cracked the code to get research ideas from the lab into product or out into the industry very effectively. Many research labs have a hard time with this, but I think Intel Labs is really good at it. We have had a lot of impact as a relatively small organization within the much larger Intel. In addition to VT, Intel Labs is the place where other technologies like Thunderbolt and Silicon Photonics originated. Leading an organization that is able to consistently deliver and transfer those technologies to Intel’s business but also outside of Intel to make things happen in the ecosystem is something that I feel really good about.”