Direct Investments

The systems behind the supercomputers

From climate modelling to astrophysics, and financial modelling to engine design, supercomputers advance knowledge and deliver results for industry. Allinea is the complete tool for software development, with everything needed for high performance. Allinea is grateful to be supported by Mercia.

Almost every aspect of science and industry is affected by high performance computing. Whilst historically early progress in the field can be traced back to its use in sciences such as nuclear physics or numerical weather prediction (both of which have massive computational requirements), today its applications are far more widespread.

Changes over the last two decades have widened its impact substantially, with the IDC (International Data Corporation) calculating the high performance computing server market to be worth around $10bn annually.

What is more, supercomputers have become a commodity and no longer rely exclusively on bespoke technology. Instead, data centres in high performance computing are now filled by racks of high-end Linux servers – or clusters – capable of doing amazing things, from exploring the cosmos, to processing seismic data for oil exploration, to computing stresses on aircraft components, to optimising the performance of Formula 1 cars.

Developing and deploying the software that uses this kind of system is a complex task. Furthermore, the cost in terms of developer time and system resources in locating and fixing inefficient or incorrect code is substantial, but the results enabled by software in high performance computing are fundamental to the continued success of corporations and institutions that adopt it, on both a national and international scale. These can include major oil and gas corporations, top ranking universities, and large research facilities.

Effective tools are therefore needed to enable software developers to solve defects and improve the performance of these complex codes, which often run across hundreds or even thousands of servers simultaneously. A quick response can help the owners of such systems discover performance anomalies and opportunities in their running applications.

However, in order to do this effectively, these tools must be adaptable and flexible. Why? Because technology doesn’t sit still in the field! Supercomputers continue to grow in size, which exacerbates the software challenges. Energy efficiency considerations are leading to new technology for processors, memory and storage, which will have an impact on the applications.

Ultimately, staying ahead of the curb in this field is vital, and working in close partnership with those using high performance computing will mean that system tools are able to face new challenges and keep up with the changing landscape of computational requirements.