How HPC research into 'miracle material' will drive commercial success

05/02/13

Biagio Lucini is a professor of Physics at Swansea University and a Royal Society University Research Fellow.  He’s using supercomputers to carry out world-class research within the field of theoretical particle physics to develop a better understanding of the newly discovered 'miracle material' Graphene. In collaboration with a Swansea based business, the results are expected to have big benefits for industry in Wales.

Here Biagio gives some background to this ground-breaking research:

How would you describe your work?

Since an unsolved or unproved theory is worth nothing in this game, we look to find solutions.  We take a theory (or a problem), turn it into a mathematical equation, and then look to find a numerical solution.  Often, it is not possible to find a solution with a pen and paper, so we use high performance computers (supercomputers) to carry out millions of calculations very quickly and from these we arrive at a prediction which we can then prove or disprove through further experimentation.

For instance, much of the current research performed in my group is centred on the newly discovered material Graphene, a sheet of graphite one atom thick that’s the strongest material discovered to date and conducts electricity at an incredible speed.  Our work involves characterising Graphene from its theoretical properties.  The material has enormous potential but is not in use on a large scale as of yet, as we don’t know enough about its characteristics.  Using supercomputers we will be able to develop an understanding of its theoretical properties using numerical simulations.

How will your research benefit Wales’ economy and society?

By developing a good understanding of the theoretical properties of Graphene, we will put businesses in Wales in a great position to take advantage of our results in their product development, building stronger products at a cheaper cost.   

It’s hard to quantify the longer term impacts to this ‘pure’ research - the first people studying electrons weren’t thinking about light bulbs or other technology like computers when they first looked at the subject; this came as a longer term result. Likewise, I’m certain there will be many longer term impacts to our line of research. 

In the short to medium term there will be a number of industrial applications the results of our research can be applied to.  A business has been sent up in Swansea already with the intention of commercialising our results in their forthcoming product developments.   

We are also contributing to the HPC sector itself.  As leaders within our field, we are developing new techniques for analysing physics data on a supercomputer.  These techniques will be available for businesses needing to analyse their data in a similar way.      

How important is HPC to your work? What benefits does it have?

The two fundamental prerequisites of my research are speed of calculation and speed at which you access large data and datasets, and therefore HPC is essential.

I wouldn’t be doing what I am doing without HPC.  It’s unthinkable to do our work without it – it simply wouldn’t be competitive.  Imagine running a factory with technology that is 100 years old, because that’s the gap between a work station and a high performance machine.  If you run a milk factory like they were running at the beginning of the last century they wouldn’t be as efficient as if they used modern technology.  That’s basically the difference for me, between using HPC and not using HPC.

Is this the first time you’ve used High Performance Computers?

No, it’s not – I’ve been using what are now referred to as High Performance Computers or supercomputers since 1996.  The first machine I used was a very early prototype supercomputer built by physicists for the sole purpose of studying physics.

How do the HPC Wales systems compare to the other systems you have used?

We are using a few different supercomputers at present.  In comparison to most of them  HPC Wales offers double the performance in terms of speed.  I didn’t expect the level of performance I got from the HPC Wales system and can say in all honesty it has been the most professionally set up system that I have used so far. It’s really a pleasure to use, because it has taken very little time; that’s not always the case, even if you are experienced, because every system is different.  With this system I felt at home from the first few minutes. 

Following a few emails between myself and the technical team, I had uploaded my data and was running my code within 1 hour of being set up on the system. 

What inspired you to do this work?

Having studied it at undergraduate level, theoretical particle physics was the area I decided to specialise in.  Along the way I discovered that using the techniques within this field you could explore all sorts of theories within material science, looking at superconductors and Graphene for instance.  

In terms of the computing element to my vocation, I was computer phobic at the beginning, scared to even switch on a computer!  I’m talking about the early 90s when there were only a few computers about, largely confined to the Universities.  I saw this big grey screen with no colour which I found quite intimidating.  There was a physics theory that I really liked and wanted to solve but was told I must use a computer to achieve this.  ‘Isn’t there any way around it?’ I protested. No, there wasn’t, was the response! I liked the problem so much I set my assistants up on the computer to start work and before I knew it I was hooked myself. 

Are there lots of theoretical physicists using HPC at the moment?

The community of my direct/indirect competitors stands at about 500 researchers at present.  Some are collaborators but most are competitors.  It is a very competitive field where three things count: your knowledge of physics, your ability to write highly performing codes, and the resources you have at your disposal at a given time.  I could be the best researcher within this field but if nobody gives me a HPC system to run codes, there’s nothing I could do.

What do you most enjoy about your work?

I enjoy discovering new things, that’s my nature.  I also enjoy working with high performance computers, the techniques involved and the writing of fast and efficient programs and watching them scale on a large number of cores – HPC is now my world!

What are your hopes for the future?

I would like to see HPC Wales growing and my involvement within it increasing.  I’d love to think if we met again in 2 years’ time I would be able to say most of my computer power came from HPC Wales. 

In terms of my research, it is a continually evolving field where new breakthroughs come roughly every other year.  You have to be quick in understanding them and getting on board. 

The line of research we are following at the moment started three years ago and I was fortunate enough to be amongst the early adopters of this technique which has given way to all the impacts and benefits we have seen so far.  It’s key to remain alert to what’s going on around us. 

What advice would you give to researchers/businesses?

It is less scary than it looks at first sight.  For those with basic computing experience you could go from using a desktop to a HPC system in next to no time.  We are talking about weeks here; in a few weeks you should be able to port your code and be well on your way.

HPC is not the solution to every problem we have.  There are problems that cannot be solved with a HPC system or, if they could, it may be highly inefficient to solve them in this way.  You first need to assess whether your problem is typical and can be bought onto a HPC system proficiently or not. Unfortunately there is a lot of confusion in the outside world about what HPC is and what it is not and that is why seeking the right advice is essential.

With the right information and the right advice it should not take long and you should not be scared to do it.