Chuck Thacker

2014 Career Achievement

Microsoft Technical Fellow Chuck Thacker is a self-­‐avowed “hardware guy.” In a career spanning more than four decades, his accomplishments have included seminal technological developments, ranging from early personal computers to the laser printer and Ethernet to, more recently, the Tablet PC.

There’s another sense in which Thacker is a hardware guy: he is a frequent recipient of trophies, plaques, and other testimonials to his contributions to the field. The 2014 TCN Career Achievement award is the latest piece of hardware to grace his trophy case, taking a place of pride alongside the prestigious A.M. Turing Award, among others.

However, as we shall see, hardware in its various forms is only part of Thacker’s story. He is first and foremost a research guy. The lab at Microsoft Research Silicon Valley, where he now works, is the latest in a long line of industrial labs that he helped to bootstrap, labs that have kept him on the cusp of innovation throughout his career.

It is perhaps ironic that Thacker’s accomplishments have taken place in a field he initially had no intention of pursing. After receiving his undergraduate degree from the University of California, Berkeley, in 1967, Thacker took a short-­‐term job with the university’s recently formed Project Genie. The idea was to earn money to finance a graduate degree in physics. An added benefit was that the underlying physics of the project might prove useful to his intended academic pursuit.

Sadly—or so it seemed—the anticipated concentration on physics was not to be. “Instead, I went to work for this computer project,” Thacker says. To understand his disappointment, it’s important to realize that at this time computers were, frankly, rather dull. “It wasn’t really very interesting,” he explains. “In those days what you did was enter a FORTRAN program on a deck of punch cards, and you submitted it to a machine, and then you came back a few hours later and got your printout. Boring.”

Fortunately, Project Genie proved to be anything but dull. The research team immediately put its efforts into building one of the earliest time-­‐sharing machines, able to serve simultaneous users. “That was interesting because it meant you could program interactively—develop the program and run it right then and there,” says Thacker. The project later morphed into a private start-­‐up, the Berkeley Computer Corporation (BCC). It produced a computer called the BCC 500, with Thacker responsible for its room-­‐size memory and processor.

Unable to find sufficient venture funding, BCC eventually dissolved. But the bulk of the team was soon picked up by Xerox and put to work in a newly founded computer science lab at the company’s Palo Alto Research Center (PARC). Their first project was the Multiple Access Xerox Computer (MAXC), a computer built to facilitate their work. (Thacker notes that it was standard practice at the time for a new lab to build a computer—not as a commercial product, but as a research tool. “We wanted the output,” he explains.)

Looking to map out the next project, Thacker considered some of the revolutionary thinking that had been floating around the lab: namely, that computers were really about interactions with people—in short, they were communication devices. He began musing on how to build a simple computer for office use that would be cheaper than the era’s mini-­‐computers that cost tens of thousands to hundreds of thousands of dollars.

The result was the Alto, which has been described as the first computer system that would be recognized today as a personal computer. In fact, Thacker vividly recalls drafting an internal memo with the title, “Alto: a personal computer with micro-­‐parallel processing.” He notes that while earlier single-­‐user machines existed—notably the LINC, or Laboratory Instrument Computer—the Alto was really the first to offer significant capacity.

Despite this success, things did not continue smoothly at PARC. Looking back, Thacker concedes that he and his colleagues essentially spent a lot of Xerox’s money emulating the future. “The fundamental problem was that Xerox didn’t understand what they had,” he explains. “They were an office-­‐system company, a copier company. They were not a computer company. We could never get the traction needed to make a success of personal computing.”

In 1983, while all this was going on, Thacker interviewed at Microsoft. He recalls it as an interesting experience, but one that left him thinking the two weren’t right for each other. “It was quite clear that Microsoft was a software company,” he says. “And I was not a software person.”

During that visit, Thacker had lunch with Bill Gates, during which the company founder described his vision of having the words “copyright Microsoft” appear on every computer screen in the world. Thacker politely suggested he consider something more realistic. It did not go over well.

Dissatisfied at Xerox, Thacker and his colleagues began looking for an alternative, and accepted an offer from Digital Equipment Corporation (DEC) to set up a lab that would work on distributed computing. Or so DEC thought. “We said no, we want personal distributed computing,” says Thacker. Their first machine was the Firefly multiprocessor workstation, followed by a system to use DEC's new Alpha chip. “We were actually able to help DEC more than we could help Xerox,” says Thacker. “Because DEC was a computer company.”

Thacker spent 14 years at DEC, turning down a job offer from Microsoft Research along the way. But by 1997, things having “gone wonky,” he was looking at alternatives. On the one hand, he was in discussion with Microsoft Research about the possibility of establishing a lab in the Bay Area. At the same time, with his two adult daughters now out of the house, he and his wife, Karen, were drawn to the idea of an extended sabbatical in Europe.

Serendipitously, Microsoft Research was in the process of establishing a research lab at England’s University of Cambridge, headed by Pro-­‐Vice-­‐Chancellor Roger Needham. Hearing of this, Thacker sent a note of congratulations to Needham, a long-­‐time friend. This led to a surprising phone call from Microsoft CTO Nathan Myhrvold. As Thacker recalls, “Nathan said, we are going to start an industrial lab, and you’ve done three of them. So how would you like to go to Cambridge for two years and help Roger set up this lab?”

Reasoning that the time had finally come, Thacker joined Microsoft. “It seemed pretty clear that although Microsoft was a software company, it was also a computing company, and as such it would evolve,” he explains. “One of the things you can say about Microsoft is that they really do turn the ship rapidly when needed.”

According to Thacker, the two years at the University of Cambridge went quickly. His research area was electronic books, although he admits to spending most of his time in an administrative role, “madly trying to hire people.”

Returning to the U.S., Thacker joined a newly formed Microsoft group also working on electronic books. This would eventually evolve into the Tablet PC, with Thacker spending two years on a working prototype. “The machine was a slate,” he explains. “It looked like an iPad, except it had a stylus, and you could do handwriting recognition with it. It worked pretty well.”

Thacker’s 2009 A.M. Turing Award citation references the Tablet PC, along with its revolutionary stylus. Erroneously referred to by some as the “radio pen,” the stylus uses a modulated magnetic field to detect the user’s intent. The functionality is vastly superior to earlier pressure-­‐sensing systems.

The Apple iPad debuted shortly thereafter and of course had no stylus. Some argued this made the iPad superior to the Tablet PC, but Thacker doesn’t agree. “It depends what you want a tablet for,” he explains. “If you want it for information consumption, [the iPad] is probably OK. But if you want to do information creation and you don’t have a keyboard, the stylus [for the Tablet PC] is the only game in town.”

After a brief foray into an educational tablet for kids, Thacker returned to Microsoft Research, joining its newly established lab in Silicon Valley. Thacker remains there today, with no intention of stepping down anytime soon (70 being the new 60, after all).

Keeping him going are the numerous collaborations that make research work interesting. He points to a collaboration he facilitated between Microsoft Research and Berkeley, in which Microsoft and partner engineering firm Celestica provided hardware to support an inter-­‐university project called RAMP (Research Accelerator for Multiple Processors). The successful project spun out a company called BEEcube, which specializes in advanced system-­‐level reconfigurable platforms.

Another interesting collaboration was Kinect. “Kinect was a combination of work done by the Xbox group, but also by Microsoft Research,” explains Thacker, noting that much of the device’s image-­‐tracking capability came out of MSR. “If we could do a few more of those, that would be pretty good. So I’ve been thinking about those opportunities.”

The potential capabilities of devices are another thing keeping this hardware guy engaged. Thacker laments that people tend not to think about what they might be able to do with a special class of device because they simply can’t imagine it. To that end, he has launched a contest among his colleagues, asking them to describe a device he might want.

When Thacker ran the concept by his biologist daughter, she suggested a small box that could be carried on nature hikes, capable of doing a DNA bar code of samples found along the way. “Curiously enough, there are some people in our lab working on the software end of something like that,” he says. “You can basically build a laboratory to do gene sequencing on a chip these days. So that might be a possible gadget.”

Thacker has certainly come a long way since taking that first job in a seemingly dull field. And he strongly disagrees with those who lament that computing has become boring. “New physics is being discovered every day, and many of those discoveries will surely have a significant impact on computing,” he concludes. “So the journey has just begun.”