As part of the steady optimization of energy consumption and performance of cluster systems, advancements in HPC environments are leading to the integration of coprocessors and hardware accelerators.
As such, our range of services includes manufacturer-specific solutions from NVIDIA Tesla and Quadro GPUs, AMD FirePro visualization processors and Intel Xeon Phi coprocessors, to NVIDIA GeForce general purpose GPUs (GPGPUs). Based on different server configurations, we equip groups of compute nodes or complete HPC clusters with one or multiple accelerators.
Our expertise in the selection and combination of GPUs is based on, among other things, many years of experience in the installation and operation of HPC clusters. In addition to current and future processor architectures, our in-house Benchmark Center keeps different coprocessors ready, in order to make the optimal choice regarding the selection of the most efficient hardware accelerators for any custom application.