High End Compute has had four successful commissions with Dr. Topping, with some further information below
- via U/Manchester "eco innovation" voucher, HEC was commissioned as a local SME to evaluate current and future technologies and their environmental footprint
- provision, at short notice (a month), of Intro to GPU Programming training to EAO DTP researchers (Spring 2017)
- provision of "Introduction to Parallel Programming" training/hands-on support to NERC researchers (Autumn2016)
- developing prototype parallel Raspberry Pi cluster, with outreach materials for secondary schoolchildren (Spring2016)
Developing prototype parallel Raspberry Pi cluster, with outreach materials for secondary schoolchildren
The Background of Dave's Requirements
Dave's current research uses computer models and a combination of mathematics, physics and chemistry, to understand how tiny atmospheric particles influence our climate and air quality. The atmosphere is full of these aerosol particles. Without them, there would be no clouds, since water vapour needs something to cling to. Discovering how these aerosol particles behave is important in understanding climate change, as well as providing information about air quality. Mathematical modelling helps us to understand the processes that affect these particles, how they respond under different conditions and how this leads to environmental effects. Making use of computational resource is very important as our science progresses. In this activity, learning how to accelerate calculations can be applied across so many disciplines, including my own. It is very important that emerging generations better understand how to exploit computer resources available to solve the problems of today and tomorrow.
How HEC Helped
Dave commissioned HEC to produce a mechanism for building a cluster out of Raspberry Pi units, a cheap LAN box and some USB power hubs. And to develop a package to take to schools in the U.K. to explain the relevance of parallel processing. The day's event takes the children from discussing computing power from the 1960s to the modern day, the role of prime numbers in cryptography, and hands-on exercises in building a working parallel RPi cluster on which they will develop and run parallel Python code using MPI.
Working with HEC has enabled me to expand into areas that would otherwise be unavailable. This has resulted in me taking my scientific remit across many areas, positively feeding back to existing and emerging challenges. Not only has Michael delivered technically, from research grade work through to outreach activities with schools, but also offered guidance on emerging roadmaps and where best to place myself to exploit opportunities. This should be the future of outsourcing technical expertise, and I would highly recommend working with HEC on all compute related matters