NAG’s Cloud and HPC Services are completely agnostic with respect to hardware and software technologies and cloud vendors. We work with all major cloud providers and hardware manufacturers on a wide range of programming languages and operating systems.
Wherever you are in your HPC journey, from optimizing and scaling your C++ or Python analysis, to a multi-million dollar, multi-application migration, from on-premise to Cloud HPC, NAG has the expertise to ensure your success.
NAG’s Cloud HPC Migration Service provides support as you
- Choose Your Cloud – we work and partner with Google, AWS and Microsoft, NAG knows the strengths of the top Cloud Service Providers and can help you select the best for your needs
- Navigate the Transition – NAG helps you understand which factors will limit performance and increase cost in the cloud (I/O, CPU, memory, interconnect)
- Optimize for Performance and Cost – NAG optimizes both your applications and the cloud infrastructure they can run on. Tuning and modernizing existing code, choosing right resources – all to reduce overall cost and/or increase performance.
HPC systems enable quicker results at lower cost. NAG’s Software Optimization and Code Modernization services help maximize HPC investment by directly improving analysis and simulation pipelines across several dimensions.
- Hardware Choice: Can GPUs or other accelerators improve your speed-to-solution or should you focus on optimizing for the CPU?
- Algorithm development and selection: Our consultants have extensive experience in the development and implementation of novel numerical algorithms, ensuring that you get the right answer quickly.
- Efficient Parallelization: HPC Application profiling to ensure your MPI, OpenMP, Spark or GPU code is using resources efficiently.
Faster Code with GPUs and Other Hardware Accelerators
In recent years there has been a Cambrian explosion of accelerator technologies that promise to run software more quickly than ever before. Which of these technologies should you choose for your workload and, once chosen, how can you ensure that you are getting the most out of the hardware?
Watch NAG Discovery Manager, Reid Atcheson, reveal how NAG helped a client achieve performance gains on an NEC C++ against traditional CPUs.
- Focused performance studies, for example a particular application benchmarked across a range of HPC systems
- A managed technology evaluation service, where we collaborate to undertake continual performance evaluations of new software and hardware technologies as they emerge and show potential for your business
- Advise on the role of benchmarking in procurement, application planning, and acceptance testing
NAG provides training courses in a range of technologies that can be delivered either at NAG offices or at client premises. Courses can be tailored to suit a particular environment and be targeted at novice, intermediate or advanced levels. Subjects include:
- High Performance Python
- GPU Programming with CUDA and OpenCL
- OpenMP
- MPI
- Fortran Modernization
Industry NAG HPC Client
