A two-way street

This was the second time that New Scientist magazine has run its flagship event, and from all reports it was bigger and better than the previous one. Along with the Big Bang Fair in Birmingham, this was one of the largest events that EPCC and the ARCHER Outreach team have attended.

There were five zones dedicated to different fields of study at the event: Humans, Engineering, Technology, Earth and Cosmos. Although ARCHER can support science in any of these areas, the stand was located in the Technology zone.

While we have attended the Big Bang Fair before, this event was different in some respects, the most obvious being that the age range, slightly higher in comparison, having many more adults and college/university students in attendance. On the Saturday and Sunday of the event this was more evident. And the visitors were not just those interested in possibly working in science and engineering but included many individuals who already work in the fields, who were along to see what was happening and talk to people about their work and pick up or provide ideas.

While large stages at the rear of the hall had frequent large audience talks, we found that the more interactive nature of our activities and the related small group or individual talks suited our style of engagement.

The team that was present for the event covered a lot of aspects of the work, which EPCC carries out both in relation to ARCHER and other areas of its business. We had ARCHER staff who help support the users and carry out ARCHER technical work (Jo, Gordon and Weronika), staff who are involved in the Software Sustainability Institute who could talk about a range of topics concerning software sustainability (Mario), staff, including our MSc Programmes Officer (Ben), who are involved in different areas of our training and education activities.

We possibly had a member of staff for almost any type of question relating to ARCHER and EPCC present. Common questions included:

• What actually is a supercomputer?
• What do you do with it?
• Do we really need one?
• How big is it?
• What type of things does EPCC do?
• Can we hire time on the systems EPCC runs?
• How do we study or train with EPCC?

Our exhibits were taken from our core range, which EPCC blog readers or those who have visited us at different events across the UK will recognise.

The bean-bag sorting activity made another appearance, letting the visitors unleash their competitive natures while allowing us to explain about how scaling and communications can affect the performance improvement achieved in parallel systems. As noted by Mario, often it was not the obvious person who was competitive at this activity. While a simple task, this illustration of the concept can be relatable to many audiences and across the event we had a range of audiences who were willing to participate. For more information on this activity please visit our activities page.

We frequently bring old hardware from different HPC systems to events, and for this event we had brought three boards from earlier eras of systems hosted at Edinburgh. As noted by some of our staff, we had a little bit of board envy as the Met Office had an XC30 board (ARCHER is composed of the XC30) and we didn’t. A new twist was that people could have their picture taken with the boards, for the selfie aficionado this proved to be popular though others weren’t so keen. The rate of advancement and change over the last 30 years alone was of great interest to audience, and spurred many conversations about what is next for HPC and general computers.

Wee Archie Blue was our final activity (its counterpart Wee Archie Green was on its way back from an event in China). This functional Raspberry Pi model supercomputer has proven highly popular with our versions having travelled to many places around the UK and the world, so much so that other places are building their own versions. While the unit is a collection of flashing lights, which will distract and attract almost everyone, without the software demonstrations it would just be a pretty sculpture. At this event, we chose to run the Computational Fluid Dynamics (CFD) Wing design demonstration. This version included the original simulation by Gordon and all of the tweaks to the interface a visiting intern made earlier in the year.

The software allows the user to design the cross-section of a wing and the cluster will run a CFD simulation using the input design. Then we have two test outcomes, can the wing be used to take off and how far could it fly on one tank of fuel. This allows the power (even at this small scale) of HPC and simulation to be highlighted, with many visitors getting to this conclusion without our input, just from using the active demonstration.

Amongst our activities, we had examples of the research carried out using ARCHER, with postcards of generated images and short summaries.

Many visitors found the images interesting but the application of the ideas behind them even more so. Their imaginations captured they could move beyond the short explanation we provided to see how something similar could be done in another field or how it could be used to improve something they knew about.

Perhaps this is one of the most rewarding elements of this style of event and interaction, seeing that with a little prompting and some examples, people can even without a background in computing or science see the value and application of this work. To see them inspired even just to learn more and to be willing to challenge us on what could be possible is a great thing.

This blog entry is called ‘A two-way street’ for a reason, and the public listening and reacting to us is one part of it, but the other part is that we can learn from our audience. One of the most interesting conversations was with a couple of medical physicists who were enthusiastic about the technology and current research, but they brought home how the advances in technology help at a more immediate level. They discussed how treatment calculations, particularly radiations treatments, over their careers have gone from requiring overnight calculation on a machine to being done in minutes if not seconds. While we may have been aware of this in broad concepts, having people working in the field explain the changes to their careers and patients to us was enlightening. This was just one example of such an exchange, others involving things like medical tests or product design were had, but it shows we can all learn from others as they can learn from us if people are willing to listen.