Skip to main content

BeginnerHPCResearch

Breaking down walls

10 November 2014

The fall of the Berlin Wall happened 25 years ago and I thought a symbolic look at how we try and break our own walls would be quite appropriate.

Creating the creative economy

The Creative Industries has gained a lot of traction recently within the UK with much emphasis on TV and film having been made in the UK and more locally within Wales. Historically the arts have not used supercomputers – except for the special effects and computer animations. We are trying to reach out to the arts and find ways we can help researchers explore new avenues of research using supercomputers.

Examples in the USA where which has happened is in iChass. This shows projects which have allowed the arts, humanities and social sciences to use supercomputers to explore new research opportunities.

A few projects have started in Cardiff University such as Lost Visions and COSMOS which are allowing avenues of research which without supercomputers would not be possible. The issue is making researchers (who are not traditional users of supercomputers) aware of what supercomputers can do and how it can interface with their work is the wall we are having to tear down.

To give or teach

give someone a fish and you feed them for a day; teach someone to fish and you feed them for a lifetime

To relate to the famous saying about fishing, the other wall we should want to break down is to share the knowledge and expertise. This is especially true in emerging economies where they should not be dependent on traditional sources of expertise and instead develop their own communities. A great example of this is Python Namibia 2015 which is being supported by Cardiff University. Supporting communities to grow will allow fresh approaches to problems and evolve according to the local requirement.

Locally ARCCA is starting to engage with teaching rather than training registered users. Supercomputers are becoming a critical resource within disciplines and students should not be hidden from the power of them. For example the biosciences now require supercomputers to analyse genes and it is now possible to sequence genes for the fraction of the cost compared to a decade ago and therefore becoming common-place within Universities.

“Looking for freedom”

Underpinning all this talk of reaching out to new disciplines and arenas, much of this would be easier if people used Open Software. Recently we had users from a non-traditional user of supercomputer wanting to use a well-known commercial math package on our system. It resulted in weeks of frustration due to dependencies on additional commercial packages. Could the work have been better placed using something like Python – possibly. The point is that if researchers choose to use proprietary/commercial software then it becomes harder to replicate the results and the consequences (possibly driven from funding councils) should be advertised more. Open Data initiatives are trying to free up data but if it requires access to software which is not open then can the results be replicated. Supporting open software is not just to save money – but also to allow greater collaboration between people and help break down those walls.

Computers without borders

Unlike the Berlin Wall – walls are rarely broken down overnight but we hope to expand the usage of supercomputers across all the different fields of research happening at Cardiff and ask researchers to contact us with any idea which might be novel in their field.