NAG in High Performance Computing (HPC)

Computation is a strategic business activity within many companies – from market & source data processing to product development to business processes and logistics. Increasingly companies are turning to High Performance Computing (HPC) to accelerate their computation.

"Study after study has showed that virtually all businesses – large and small – that adopt HPC consider it indispensible for their ability to compete and survive. The reason is simple: HPC is a proven game changing technology."
“Advance. Benchmarking Industrial use of High Performance Computing for Innovation,” USA Council on Competitiveness in partnership with IDC, 2008.

Why Use NAG

Organizations can achieve an enhanced commercial advantage from computation by partnering with NAG’s High Performance Computing (HPC) Consulting Service. NAG experts support the whole business lifecycle including requirements review, HPC strategy, resource and service delivery planning and acquisition, and parallel software engineering support.

Leading Computational Expertise

NAG has one of the largest collections of computational expertise of any organization. NAG’s strength is the breadth and depth of computational expertise available, with our staff possessing experience as HPC end-users (across industry, defense and academia), as HPC service providers, and as HPC software developers.

In addition to our ‘supercomputer centre class’ large scale distributed parallel programming expertise, NAG’s nearly 40 years of numerical software product development core (with strong business on mid-range HPC platforms) means we have strong established experience in multi-threaded programming which is increasingly important in all scales of parallel applications.

NAG has a consistent track record of consulting contracts from hardware vendors, software developers, the financial services sector, and more. For example, in 2006 NAG won a competitive tender to secure a 6½ year contract to provide over 120 person-years of CSE support for HECToR - the UK’s £113M national academic supercomputing service from 2007-2013.

Computational Science & Engineering (CSE)

CSE - the engineering and science of computation - is about optimally exploiting the computational facilities (software and hardware) used in your business. HPC Consulting from NAG offers three strong return on investment opportunities from CSE: extending application lifetime through parallelism, extending application reach through scalability, and enabling application innovation through algorithms.

Parallelism	Scalability	Algorithms
Incorporating parallelism in your software extends the lifetime of the application.	Engineering your software for scalability extends the reach of your application.	New and improved algorithms enable innovation in your application.
Computation is undergoing fundamental change across the industry: future processors will not be faster – they will be better at parallelism - whether multicore processors or heterogeneous computer clusters with GPU’s and other accelerators.	Processing larger datasets, including more complex physics, exploring longer timespans and larger parameter spaces, etc - the only cost effective way to access this increased demand for memory and computational power is to use more servers and processors.	More efficient use of memory, more resilient computing, inclusion of new physics, and more. In some cases, new algorithms and code re-factoring can improve application performance and scope as much as faster computers and bigger memory.
Use NAG's CSE expertise to ensure your software/applications evolve and avoid the very real prospect of legacy (non-parallel) software running relatively slower with future hardware.	Use NAG's CSE expertise to make your software scalable and optimised for the increased scale and complexity of hardware needed to deliver results in time.	Use NAG's CSE expertise to take your applications to their next generation of computation with new computational features implemented in future-looking parallel algorithms.

© Numerical Algorithms Group

Visit NAG on the web at:

www.nag.co.uk (Europe and ROW)
www.nag.com (North America)
www.nag-j.co.jp (Japan)

http://nag.com/hpc/index.asp