Skip to main content

A life of freedom

Colleagues of Larry Smarr believe that, had he chosen differently, he would have been as successful in politics as he has been in science. His eloquence and charm, and his clear tactical thinking, are as strong as his talents in theoretical physics. He has brought hundreds, if not thousands, of scientists together for huge projects and persuaded funding agencies of their national importance. Oh, and as an aside, he pulled off the key political stroke that made the Internet we know today possible.

 

All the time he was just following his own interests, and went where they took him. "Nobody has ever been able to tell me what to do," he says. "I believe I am at most use to society when left to do what I want." While many of those he worked with went on to make fortunes in the Internet boom, Smarr believes he has gained a greater reward for his years of public service - freedom.

 

Smarr is best known as founder of the National Centre for Supercomputing Applications at the University of Illinois. While there, he was instrumental in establishing the NSFNet. Although his original idea was just to connect scientists to supercomputers, it soon became the backbone of the Internet. Smarr, who says he is descended from Scots-Irish pioneers who first colonised the West, is now seeking fresh challenges as founder of the California Institute for Telecommunications and Information Technology. He is hoping to create a new Silicon Valley around the La Jolla Campus in San Diego, while continuing to push the frontiers of collaboration. He says: "What I am looking to do is to bring together teams from different disciplines to form multi-cellular organisms."

 

Forest Baskett, former Professor of Computer Science at Stanford University, which spun out Sun Microsystems, Silicon Graphics and MIPS, is one of those who admires Smarr's political skills. He says: "Larry is a delightful person and as you can imagine for someone with so many accomplishments he is full of energy all the time. He works really hard and is very effective at influencing people - rallying the troops. He probably would make a great politician. People often say he is very political because he is so effective at drumming up support both in terms of money and people.

 

"He has certainly been a major force in promoting distributed models of computing. It is leadership thinking with a lot of follow-through. In terms of the Internet, he was able to promote the NSFNet as a way of allowing academics to gain access to the supercomputing resources they wanted. But in a way this was a bit of a ruse to escape the attention of the traditional service providers by persuading them that it would never amount to anything. If you think about it that way, it really was startling what the result was."

 

Smarr grew up in Columbia, Missouri, the son of fourth-generation horticulturalists. The garden of the home he grew up in was a tourist attraction in the State because of its flowers. But what influenced Smarr more were books about maths and science left in the house by an uncle. While still at high school he was able to attend classes at the nearby University of Missouri where he soon found his metier of theoretical physics. Although he has never liked being told what to do, he did take his father's advice and took a course in computing. Straight away he found that he loved poring through reams of data and that computers were the tools he needed to follow his interests. For his PhD at the University of Texas, Austin, he applied numerical methods to solving general relativity equations for the collision of two black holes, and launched an academic career in theoretical astrophysics.

 

After several years in post-doctoral research, he settled down to his first faculty position at the University of Illinois. One of Smarr's frustrations was that the computers available to him were relatively slow and he sought out access to faster machines to run his numerical solutions. He found one of the emerging supercomputers at the Lawrence Livermore Laboratory but because it was being used in top-secret military research, he had to get a security clearance and physically travel to the laboratory in California to use it. This frustration led directly to the NSFNet. In 1983, he lobbied other scientists who wanted online access to supercomputers and the National Science Foundation to build a high-speed network to connect scientists in universities to the machines in military laboratories.

 

He says: "As soon as I mentioned a network, the service providers like AT&T, Sprint and MCI, who sat on the National Science Foundation, reacted against it because building networks was their business. They asked me how many people would be connected to it and I said a few hundred and they soon lost interest. We constructed a backbone network, running at an amazing 56kbaud, linking five centres one of which was the University of Illinois where I was able to found the National Centre for Supercomputing Research in 1985. The NSF insisted that the network used an open standard and decided to use TCP/IP. This was amazing at the time because the networking committee was dominated by physicists who mostly used DECNet. Later the Networking committee decreed that all future NSF-funded networks would have to use TCP/IP and gradually more networks were added to the backbone. Later colleges started ripping up their quads and putting in their own networks and connecting them through the backbone." The rest, as they say, is history.

 

The NCSA was keen to improve the interface for supercomputers so it first developed Telnet. When Tim Berners-Lee invented the World Wide Web a young research student called Marc Andreessen pulled all the access programmes together into the Mosaic browser, which could display images and text on the same page.

 

Smarr is far from envious of colleagues who have made fortunes from the Internet. He says: "The trouble with having a lot of money is that it takes possession of you and these people spend all their time with tax accountants. I don't want to do that. I feel I have been treated appropriately. I have the greatest luxury, which money cannot buy, and that is being left alone to follow my dreams and have enough resources to sort out the ones that are real. The fact that I have been allowed to do what I want and have never had a boss, I would not trade for millions of dollars."

 

Last year Smarr moved from Illinois to San Diego, saying that after 15 years at the NCSA, it was time to hand over to someone else. "Some people are founders and others are maintainers. I am a founder and I felt that it would be better for someone else to take over and take it to new heights."

 

Smarr believes that it is important to show people how government-funded academic research is an important driver of the economy and he is passionate about promoting this. "Research and development cannot be done by industry, at its own admission, because industrial organisations face quarterly profit reviews and cannot invest in long term R&D. The universities are creating long-term intellectual capital and only people who have long-term tenure can afford to do this. That is what my life has been dedicated to."

 

As well as developing his own research base at San Diego, he is instrumental in the movement that will build the even higher speed networks of the future.

 

Not only is he the founder of a new Institute, he has had to found a new garden since moving to southern California. When not working, he maintained extensive gardens both in Chicago and in his weekend retreat in the country. He is starting from scratch again. "It's great, I'm learning how to grow all kinds of orchids now that I could never have dreamed of growing in Illinois."

Curriculum Vitae

1970
University of Missouri B.A. Physics
1971
University of Missouri M.S. in Physics
1972
Stanford University, M.S. in Physics
1975
University of Texas at Austin Ph.D. in Physics
1975-79
Post doctoral Research at Princeton, Yale and Cambridge Universities
1979
Faculty member, University of Illinois at Urbana-Champaign (UIUC)
1985
Founded the National Centre for Supercomputing Applications at University of Illinois and became its director
1997
Director of the National Computational Science Alliance
2000
Professor of Computer Science and Engineering in the Jacobs School of Engineering at the University of California San Diego. Founding Director of the California Institute for Telecommunications and Information Technology



Media Partners