High Performance Computing - HPC

The growth of needs in computing is an issue that affects many industries: energy, aeronautics, finance, insurance.

In this context, users need an ever increasing computing capacity in order to manage the workload, keep up with deadlines while controlling costs and ensuring the consistency of the data centre.

Several approaches allow the creation of computing solutions known as "HPC" or High Performance Computing:

• Grid computing, which means using a large number of network standard machines which are middleware-controlled, to allow the parallelisation of the execution of a computation as several distributed tasks,
• "Exotic" platforms such as GPU, which means Graphics Processing Unit, that when used instead of a classic processor, allow a significant computing acceleration thanks to its several hundred cores.

Our partner in this area is NVidia : cliquez ici pour découvrir nos offres autour de CUDA^TM

• Super computers whose internal architecture is massively parallel, of the IBM Blue Gene type.

For many years, ANEO has led many missions and very active capitalisation on these topics. This has led to technical expertise and know-how which are recognised in this field.

ANEO supports its clients through the different phases of the implementation of these performance oriented solutions:

• Preliminary studies: choice of an appropriate solution, preliminary review, ROI study, technology comparison...
• Implementation: deployment of a tool or technology, program or algorithm porting in order to take advantage of the possibilities offered by the chosen technology;
• Operation: help in setting-up and customizing the solutions chosen, drafting up of guides and best practices for the use and operation of technologies;
• Optimization: cost reduction approach. Infrastructure integration, computer network integration, computer network renewal... ROI studies and help in deploying the optimization plan.

Contact : Benjamin LOO