While the Apple iPad dominated headlines in recent weeks, #b#Vivek Pai#/b# continued with his daily life teaching courses at Princeton University, meeting with students, and checking E-mails. But few know the computer science professor is on the cusp of revolutionizing Internet usage.
For the past several years, Pai has led the team of researchers that has created HashCache, a new, more efficient data storage system that would make Internet access more affordable, particularly in developing countries and remote areas where networks are limited, unreliable, or slow. The system, which would also expand Internet usage globally, was named as one of the top 10 emerging technologies of 2009 in Technology Review, published by MIT.
“It’s a way of efficiently distributing large files to large numbers of people,” Pai says. “You can use this new piece of software and a lot of people can download information from the Internet at the same time,” Pai says. “As homeowners get higher speed connections, the demands placed on the content providers have become enormous. People can use our system to offload work from their own servers to more computers and the speed of download will be much better.”
Pai will discuss HashCache during “Taking University Inventions to Market” on Tuesday, February 16, at 5:30 p.m. at Princeton University’s Friend Center Auditorium. The presentation is part of a free panel hosted by the Keller Center in conjunction with the Jumpstart New Jersey Angel Network, IP law firm Drinker Biddle & Reath, and Princeton University’s Office of Technology.
Other panelists include #b#Katherine O’Neill#/b#, executive director of Jumpstart New Jersey; #b#Shahram Hejazi#/b#, a venture partner at BioAdvance; #b#Thomas Mc Williams#/b#, a partner at Drinker, Biddle & Reath; and #b#Laurie Tzodikov#/b#, of the Princeton University Office of Technology Licensing and Intellectual Property. #b#Ed Zschau#/b#, a Princeton professor, entrepreneur, venture capital investor, and former U.S. congressman, will moderate.
Pai began experimenting with computer science in sixth grade, when his middle school received a small shipment of Apple 2-Plus computers. Realizing there were not enough machines to accommodate the students, he persuaded his parents to buy one of the pricey computers for their home. Having a personal computer, he says, provided his first glimpse of inventing, and with a few clicks, he was hooked. He was able to write his own programs and produce “simple, silly graphics” that moved across the screen. “Just that feeling of control was amazing,” he says.
Back then, even as Pai was writing complicated software that surpassed his teachers’ knowledge, he didn’t consider himself a scientist or inventor. However, he knew he wanted to pursue a career in programming. He earned his bachelor’s degree in electrical and computer engineering from Rice University in Houston, where he grew up with his father, an electrical engineer who worked mostly in power distribution, and his mother, a data entry operator who later became an accounting clerk.
“I thought I’d end up being a programmer somewhere,” he says. “I worked as a programmer for two years and quickly found that, with just a bachelor’s degree, it was harder to move up in the ranks and get interesting assignments.”
Realizing he needed to return to school, he took a leap of faith and quit his job. “The general advice from co-workers was if you find yourself here for five years, you’re never going to go back to school,” he says.
“By then, you’ve gotten used to a higher salary and to spending more money. You’re never going to give that up to be poor for five years to go back to school and get an advanced degree.”
“Once you quit your job, you might as well go for broke,” he says. “I was young. I was single. I didn’t really have much in the way of expenses.”
Pai earned his master’s degree and doctorate in electrical and computer engineering, also from Rice. Shortly after, he joined the Princeton staff in 2000.
Today he largely focuses on server performance, operating systems, and networking, as well as security and reliability in wide-area distributed systems, including content distribution networks. His research group has developed the CoDeeN content distribution network, which receives more than 25 million requests per day. He also helped develop CoBlitz, which provides scalable large-file transfers from PlanetLab — a research project started by Larry Peterson, chair of Princeton’s computer science department, that includes a collection of hundreds of machines spread around the world — to the general public.
Pai started working on HashCache in 2004 and has since commercialized it and held trials at various companies. The name, he says, combines two tech terms: hashing, a method of scrambling information; and cache, a storage method. “It sounds cool, so it stuck,” he says.
The system, he says, is revolutionary because it can be deployed in a very low-cost box. It can also store more information from frequently visited websites on a local hard drive, thus enabling direct data access and reducing the cost of maintaining individual hard drives, which can be expensive to maintain because they require large amount of RAM space.
“We still have a research program, so we can take some of these same concepts and apply it to low-end systems in schools and libraries in the developing world where there are no high-speed network connections. It’s much easier to manage. The speed of download will be much better, and people will be able to download high-quality videos at home as bandwidth increases.”
Though the team is not targeting businesses just yet, “the underlying technology could be used to speed up communication between central offices and branch offices for some companies. Systems to perform this kind of compression exist now, but they get pretty pricey,” he says. “HashCache would allow us to develop one with less hardware, translating to lower cost.”