FEATURE
Tags: 

From correcting errors to building quantum computers

John Murphy profiles the Director of the Institute of Quantum Computing at Waterloo University, Ontario

It has been well known for some time that the feature-size of silicon chips will soon get to the stage where quantum mechanical effects will start to show. One solution that has been offered to the problem of continuing to increase the power of computation is to use quantum effects themselves to create a new paradigm of computing - quantum computing. It was always thought that a classical computer could efficiently simulate any computer architecture that could be imagined, but quantum computers seem to have broken this rule. Classical computers are very bad at simulating quantum computers so, if they could be built, then maybe they would be a much more powerful model of computing. However, it is a big 'if'.

These machines are in their infancy, and one of the first crucial steps in their development has been proving that they are not impossible to build. Quantum computers, like classical computers, make errors, but Raymond Laflamme, with others, proved that it was possible to correct those errors without creating more errors than you correct. Once he proved that it could be built, the next problem he has tacked is actually to build something, although he admits that that could take some time.

Raymond Laflamme played a leading role in this proof while at the US Los Alamos National Laboratory. He has since returned to his native Canada to take things further, with experiments at the Institute of Quantum Computing at Waterloo University in Ontario where he is director. After a career, so far, as a theoretician he is now also chief experimentalist leading a multi-disciplinary team of physicists, mathematicians, and computer scientists, doing some fundamental research in the field.

Michele Mosca, deputy director of the Institute said: 'Ray is very down to earth and unpretentious. He is a great storyteller; he likes to spend time with his kids. He is very practical. He puts in his own hardwood floors and fixes his own car - which is a very old VW bus. Even in his scientific career, he is very connected to the real world. He does the highbrow fundamental stuff, right down to working with a wrench fixing a magnet.

'He is tremendously broad. He does collaborate with other people, but he doesn't just leave them to worry about aspects of a problem.

'He is recognised as having made an important contribution to error correction in quantum computing and the fault-tolerance threshold. There were many other people who played a part, but Ray certainly had an important role. They proved that there was a fault-tolerance threshold, and so it would be possible to build these machines. If it did not exist, then people could argue that a quantum computer could never be built.

'One of his strengths as a researcher is being able to bridge different fields and bring them together. Firstly, because he is willing to get his hands dirty and learn about them, and secondly, he respects other disciplines. He has been sympathetic to building up the mathematics and computer science expertise here, even though he is a physicist. He recognises talent in all its different forms, and is even-handed in allocating resources. He is very keen to bring together all the elements that are needed to make this work, not just his own field.'

Laflamme was born in Quebec City in 1960. His early years saw a resurgence of the French Canadian identity, and he was one of the first generation to grow up proudly expressing his culture. His father was a medical doctor and his mother a dietician. He was always interested in science at school, but not to the detriment of his interest in sports such as skiing, kayaking and skating - although he lacked the physical bulk to play the national sport of (ice) hockey. He can remember bunking off school to go skiing.

His family was very interested in culture, and took a great interest in encouraging their children in schoolwork, art, and music. He did well at school, and went to Université Laval to study physics. His lecturers were obliged to quote references in French books, but he and his fellow students soon discovered that the English versions of the same books were half the price.

While at Laval, he became interested in quantum gravity and wanted to pursue this interest in a PhD. His mentor at college suggested that he apply to Cambridge University in the UK, if he wanted to study quantum gravity. Laflamme thought someone from a less well-known university like Laval would stand very little chance, but his mentor said: 'If you don't apply you certainly will not get in,' which came like a slap in the face.

So he applied, and got in. He was accepted in a PhD programme with Professor Stephen Hawking as his supervisor. 'The person who was most surprised at this happening was me,' said Laflamme.

He found working with Hawking a rewarding experience, particularly as the communication challenges meant that Hawking did not waste time on anything that was not important. Time was, of course, Hawking's speciality. He had proposed that time changed its direction at the moment of maximum expansion of the universe and, during contraction, time would go backwards. This would, of course, have violated the second law of thermodynamics.

Laflamme was able to show that if this were the case, then the conditions during the Big Bang would have conflicted with observations of cosmic radiation. It took Laflamme six months to convince Hawking that he was wrong, and the second law survived to fight another day. He was given credit in Hawking's bestseller A Brief History of Time.

Laflamme said: 'I found it very inspiring that, as a PhD student, I could work on problems that so many great scientists like Boltzman had worked on. Stephen was incredibly grateful. He said explicitly that Don Page and I had convinced him that he was wrong. It's not bad for a PhD student.'

Having made his mark, he was offered a two-year post doc research fellowship at Peterhouse College, Cambridge, but at the same time got an offer from the University of British Columbia.

He said: 'I love skiing, so I thought I could postpone the research fellowship and go to Vancouver and come back in two years time, and I could go skiing on top of that. It turned out to be wonderful, because that is where I met my future wife.'

After two years in Vancouver, he returned to Cambridge for two years before being offered a job at Los Alamos National Laboratory. The Los Alamos job appealed, because he would be working with Wojciech Zurek, a leading scientist in astrophysics and quantum mechanics. It was also close to the Rocky Mountains, where there was very good powder skiing.

Laflamme said: 'Los Alamos was a great place to work; not only does it have good people, it has an incredible amount of resources. Travelling around the world at Los Alamos is nothing to the budget. It's also very keen on multi-disciplinary research, so it is very open for people to change the area that they work in. This allowed me to move from astrophysics and quantum mechanics to quantum information and quantum computing, which I learned a lot about from Zurek.'

Laflamme started to work on proving that it would be impossible to build quantum computers, because it would never be possible to isolate a quantum system enough to preserve its quantum properties. As luck would have it, a former colleague in Vancouver published a paper proving this point: this 'scooped' him. Laflamme was so angry that he set about proving the exact opposite. That led to quantum error correction and the fault-tolerance threshold.

Of course Laflamme realises there is a long way to go before useful machines are built, but at least he now knows it is possible to control quantum systems, so there is no fundamental reason why they should not eventually be built. There are many models for what quantum computers might look like, and Laflamme is working on using nuclear magnetic resonance. He has also worked on using linear optics.

In 2000, Mike Lazaridis, creator of the BlackBerry and founder of Research in Motion, gave $100 million to start the Perimeter Institute for Theoretical Physics at Waterloo, near Toronto. Laflamme had become well known in his field, so he was approached to become director. He wanted to carry on doing experimental work, so the University of Waterloo agreed to let him found an institute of quantum computing with Michele Mosca, who had been part of a leading team in the field at Oxford University.

It may sound like blue-sky research, but Laflamme has not found any problem getting funding. The obvious application for very high-powered future computers is in code breaking, and the US National Security Agency has taken a very strong interest. Laflamme said: 'The killer application is cryptography. The NSA sponsored my work because they wanted to know if they would have a problem, 20 or 30 years down the line, with information which they encrypt using public-key infrastructure today, that they still want to keep secret. If it can be cracked in 20 years, then they have to change their cryptography today. If we had told them it would take a year or two, all our work would have been classified.

'The NSA is very open about its sponsorship. They are an incredibly good funding agency. The paperwork is kept to the minimum; they come and visit; they ask you good questions; they are well informed. I had a biased idea before, that they were amateurs, but they are very professional and have a clear idea where they are going. They are at a level that I wish all funding agencies could be.'

The IQC has six faculty researchers and is recruiting six more this year and six more again the next year. Laflamme believes that Waterloo can become an important hub for quantum computer research, and will be able to attract the top minds from around the world that have been put off working in the US in the current climate.

Curriculum Vitae

1981-84
BSc in Physics at Universite Laval, Quebec City
1984-85
Cambridge, England, Part III of Mathematical Tripos
1985-88
Cambridge University, PhD in the Department of Applied Mathematics and Theoretical Physics (DAMTP)
1988-90
University of British Columbia, Killam post-doctoral fellow
1990-92
Cambridge University Research Fellow at Peterhouse college
1992-2001
Los Alamos National Laboratory. He arrived as a Director-funded post-doctoral fellow, became an Oppenheimer Fellow in 1994, became Technical Staff in 1997
2001-present
Founded Perimeter Institute for Theoretical Physics at the University of Waterloo; started the Institute for Quantum Computing

Feature

Robert Roe reports on developments in AI that are helping to shape the future of high performance computing technology at the International Supercomputing Conference

Feature

James Reinders is a parallel programming and HPC expert with more than 27 years’ experience working for Intel until his retirement in 2017. In this article Reinders gives his take on the use of roofline estimation as a tool for code optimisation in HPC

Feature

Sophia Ktori concludes her two-part series exploring the use of laboratory informatics software in regulated industries.

Feature

As storage technology adapts to changing HPC workloads, Robert Roe looks at the technologies that could help to enhance performance and accessibility of
storage in HPC