The universe in a desktop
The image of an astronomer is someone who spends their nights staring at the skies, these days through several million dollars worth of hardware, rather than the humble telescope. Carlos Frenk, Ogden professor of fundamental physics at the University of Durham, has been a pioneer of a different approach. Instead of just observing the footprints of the birth of the universe he has sought to recreate it as a mathematical model inside a computer.
The field of computational astronomy is now well established, but when he began in the 1980s at the University of California in Berkeley, it was bordering on heresy. Working with a small group of colleagues they started producing primitive models, which became increasingly sophisticated based on the physics and the current theories. When they started to produce results that supported some of what were then maverick theories, the whole of cosmology was turned around. When they went on to produce predictions that were supported by the latest observations these maverick theories started to become the orthodoxy.
Frenk has built up one of the top computational astronomy departments in the world. At the same time, he has achieved a high profile for himself and the field of cosmology in the mainstream media, largely because of his ebullient personality and the excitement with which he talks about his subject.
Frenk’s close friend and collaborator Simon White, director of the Max-Planck-Institut für Astrophysik, says his enthusiasm is infectious and he has a tremendous ability to communicate that, both to others in the field and to the public at large through media appearances.
He says: ‘He is a larger-than-life personality and he likes to talk. He is strongly committed to his science and his enthusiasm is infectious. He likes to say “if you don’t shoot you don’t score”. He likes to shoot the ideas out and sort them out later. He thinks that the department at Durham would collapse without him, but I think they might just survive.
‘He got in at the beginning of something that turned out to be quite productive. It was a very exciting time. A lot of the ideas that are now taken for granted were regarded as being quite way out when we started. We have seen the things we did change from being some suspect outside activity – things that serious people didn’t worry about – to being the standard story about how the universe started. He was there and his enthusiasm contributed a lot.
He keeps on firing ideas at you, but he also lets other people contribute ideas. ‘We were part of a change in the way computers were used. They used to use them just to calculate the answer to an equation that you couldn’t solve, and later used them almost experimentally to try to understand how the universe evolved; that is now a much larger part of our subject.
‘He was one of the youngest members of the team and so he started as a junior partner. In the early stages he did a lot more of the actual work. He’s an excellent communicator and he has always gone out to convince people that this is the way to go and push people to produce work that really has some impact. He is particularly good at talking on television. He is a natural communicator and he does not have to think about it too much; he just sits down and the snippets come out. Most scientists now see that as an important part of their job.’
Frenk was born in Mexico City. His father is a doctor involved in medical research, and his mother is a Spanish chemist and former concert pianist. His grandfather had escaped Nazi persecution in 1930s Germany and had helped many other Jewish refugees escape to Mexico, Canada and the US. The family was heading to Canada when Frenk’s grandmother accidentally collided with an official at the Mexican consulate in Hamburg as she came out of a bookshop. They started talking and the family decided to try Mexico instead.
His family were split between music and science, but Frenk preferred science. At school he showed some talent for basketball and played semi-professionally, but discovered he was not tall enough to turn pro. The other half of his life at school was mathematics. He went to the University of Mexico to study theoretical physics and ended up being an exceptional student, gaining the highest marks for a decade and winning a prestigious medal.
He says: ‘I only dropped two marks, and only because of one guy who said I was too arrogant. He said he was doing me a favour; he didn’t want it to go to my head.’ He started looking around for a graduate school and had been offered a place at CalTech. In 1976 he travelled to Italy, where he attended a guest lecture from a Cambridge Professor called Martin Rees – now Astronomer Royal and president of the Royal Society, Lord Rees.
Frenk says: ‘He is truly a great man and I can count the number of great men I have met on one hand. He was teaching at a summer school and I was so impressed that he made me change my mind and I then wanted to go to Cambridge. He invited me to go to Cambridge, so I changed all my plans and, luckily, got a scholarship.’
At the time his interest was split between astronomy and particle physics. Initially he applied to do the Mathematics Tripos Part 3 and he figured that, whichever end of the size scale he chose, it would be helpful to him even though it was an extremely tough course. In the end Rees’s influence persuaded him to choose astronomy, and he was taken on to do a PhD.
At this time the idea of dark matter was extremely speculative and Frenk decided that he would investigate the properties of the Milky Way for his PhD. He concluded that the galaxy was surrounded with embedded dark matter. He worked at Cambridge at a time when the historic rift between Sir Fred Hoyle at the Institute of Astronomy and Sir Martin Ryle at the Cavendish Laboratory still restricted collaboration between the two groups. He first met Simon White during this time, but White was already a postdoc and was Frenk’s unofficial supervisor.
After completing his PhD he was due to take up a position in Germany – when he got a telephone call from Marc Davis, who had just left Harvard to take up a job at UC Berkeley. Davis had completed a famous map of 2,200 galaxies and was looking for a theoretician to help interpret it. White had also moved to Berkeley and the scene was set for them to work together on a significant advance in astrophysics.
Frenk says: ‘This was the beginning of the most exciting period of my life, because we started to break new ground and explore models that were speculative at the time. There were many creative ideas that had been put forward at the time concerning dark matter and the theory of inflation and vacuum energy.
These ideas had huge implications and we had an idea for dark matter and an idea for the early state of the universe. We thought we could put these ideas together and calculate a mathematical model for the whole universe. That was almost the beginning of computational astronomy.
There had been an attempt before. ‘In those days it was so new; the idea that you could model whole galaxies was quite revolutionary. We knew how to do it, but we did not have the computer code, so we recruited George Efstathiou, who at the time was a PhD from Durham, as it happened. We would never have guessed how far we would go. One of the first conclusions we came to was that the dark matter could not be made of neutrinos. We wrote a series of papers in a few months and showed that the particle physicists were wrong.
Some theories of supersymmetry came up with the idea that the dark matter particles were “cold” (as opposed to neutrinos which were “hot”) and we could not rule this out.’ Davis, White, Efstathiou and Frenk became known as the ‘Gang of Four’ after their research started bringing in important results despite the limitations of computing power available at the time. The theories with which they started developed into what is today the current paradigm of the formation of the universe, and while Davis has been away from the scene due to health problems, the three remaining gang members have become the backbone of a massive collaboration across Europe and the US, known as the Virgo Consortium.
By 1983 Frenk had married a British woman, and they could not agree whether to live in the UK or the US. Frenk had moved to UC Santa Barbara and also spent some time at the University of Sussex in the UK, while they decided in which country to settle. The collaboration continued, with Frenk travelling a lot.
Eventually the family settled in the UK. The mid 1980s were not a good time to be looking for an academic job in astronomy in the UK. Frenk could have easily found a position in the US, but he struggled to find any opening to apply for in his adopted country.
When he got a lecturing job at the University of Durham, it was like winning the lottery. He says: ‘I think there had not been a job available in a UK university in astronomy for a decade. But we had our first child in 1984 and I had been offered a temporary SERC fellowship for five years in Cambridge. We had actually moved to Cambridge when the Durham job came up. I came home from the interview and my wife thought I had been rejected. In fact I had been offered the job, but I wanted to stay at Cambridge. My wife argued that I needed a secure job and it was the first one for 10 years – I might wait another 10 years for another one. Cambridge was an old place, an established place – and I had ‘done’ Cambridge. Durham was new for me. The physics department at the time was tiny, with no theory and the astronomy was non-existent. So this is a challenge.’
He rose to the challenge, with the full support of Richard Ellis, who recruited him to Durham. He now believes that Durham is the pre-eminent institution for astronomy in the UK and second only to the Max Plank Institute in Europe. It was not easy. Frenk struggled to get access to the computing power that he needed to advance his computational approach. He tried to borrow computers from commercial laboratories until finally he got his own MicroVAX at a cost of £40,000. He was the envy of all his colleagues.
During the late 80s and early 90s Frenk and White’s ideas were under siege. All that changed in 1993, when the CoBE satellite produced results that supported their theories. George Smoot, who shared the Nobel Prize for Physics in 2006, was a friend of both from Berkeley. Frenk is now one of the top 10 most cited astronomers in the world.
Frenk believes that the government funding bodies were adopting an approach of sharing scarce resources around to everyone, so that nobody had enough. He describes this as a recipe for mediocrity, although it is seen as fair. He contrasts it with the German approach, where fewer projects are supported, but they get all the resources that they need to succeed.
The UK attitude changed in 1994 with the announcement of the High Performance Computing Initiative, which promised lots of resources and invited bids from consortia. By this time, White was at the Max Planck Institute and the two friends joined their institutions with other computational astronomers to form the Virgo Consortium. The UK government contributed funds to build up supercomputer resources (it now runs what is called ‘The Cosmology Machine’ with 792 CPUs), but the relationship with the Max Planck Institute gave Frenk’s team access to the Max Planck Society’s world class supercomputing centre (which has 500 ultra-sparcIII processors, and the IBM Regatta system with 816 power-4 processors) in Garching.
The most famous project of the collaboration is called the ‘Millennium Simulation’, the largest N-body simulation ever carried out. It took 28 days to run and consumed a total of 343,000 hours’ worth of CPU-time. Frenk recounts being called by the system manager at Garching, who said his calculations were bringing their largest machine to its knees.
Frenk thought, at first, that he was angry – then realised he was calling to congratulate the team. The simulation resulted in some spectacular images of what the early universe might have looked like and this caught the imagination of the mainstream media. Frenk started to turn up on radio and TV, with his infectious enthusiasm and natural style proving a hit with audiences and TV producers alike.
Frenk has always regarded communication with the wider public as an important part of his job, not least because most of his work is paid for by general taxation. He also spends a lot of time on outreach work in the North East of England hoping to engage the next generation of physicists and astronomers.
In this he has found a powerful ally in the form of Sir Peter Ogden, joint founder of Computacenter. Ogden had studied physics at Durham before entering the business world. When Computacenter went public in 1998 he decided to donate his proceeds from the float to a charitable foundation supporting educational projects.
As a result of Ogden’s sponsorship Frenk’s Institute of Computational Cosmology has a bright new building, which Frenk regards as an inspiration to those who work there. He is also now Ogden professor of fundamental physics and the trust supports his outreach work.
All this contributes to a very comfortable working environment. Frenk says he has been approached by US universities, but he prefers to stay where he is. He says that, while the UK might not be a generous funder of science, it has great scientists and the relationship with the Virgo Consortium means he has access to the best of both worlds.
Carlos Frenk CV
1972-1976 BSc Theoretical Physics.
University of Mexico
1976-1977 Maths Tripos Part III
(Honours) University of Cambridge
1977-1981 PhD (Astronomy)
University of Cambridge
1981-1983 Postdoctoral Research
Fellow University of California, Berkeley
1983-1985 Postdoctoral Research
Fellow University of Sussex
1983-1985 Assistant Research
Physicist University of California, Santa
1985-1991 Lecturer in Astronomy
University of Durham
1991-1993 Reader in Physics
University of Durham
1993-2001 Professor of Astrophysics
University of Durham
2001-present Ogden Professor of
Fundanemtal Physics and Director,
Institute Computational Cosmology
University of Durham