Services
High Performance Consulting is an authorized Nvidia partner with industry-leading experience as a GPU computing solutions provider. Since 2009, HPC has helped clients leverage GPU computing to new levels of performance and power efficiency. HPC is also recognized as a driving force in the GPU computing community, presenting research results at leading technology forums and conferences.
Software development
We offer development of efficient, high-performance GPU-based computing solutions in both CUDA and OpenCL. We manage everything from parallelizing and mapping your computations to GPUs, to programming, verification and final software integration. Our client projects span a wide range of disciplines, including:
- Signal and image processing
- Medical technology
- Finance
Our experience includes both working in-house at clients on a day-to-day basis and remote work from our own offices; project needs dictate which works best. To stay on top of the rapid developments in GPU computing, we regularly attend and present at the main GPU computing conferences – see our publications and talks.
Training
Beyond software development, we offer two types of training: one at a more general managerial level and one developer focused.
GPU Computing Introduction – 2 hours
Gives engineers and technical managers a solid introduction to today’s key technologies in the GPU computing space, highlights where the technology trends are heading, and provides valuable orientation in current hardware and software. The first hour generally offers best value for technical managers and system designers, while the second hour is more directed towards software developers.
GPU Computing Developer Foundation – 2 days
Gives software developers and engineers a solid foundation in implementing the best technologies available today in the GPU computing space. Theory and concepts are interleaved with hands-on coding exercises, following the principles of problem-based learning and learning by doing. The course is based on one of the two dominant GPU computing frameworks, CUDA or OpenCL, and problems and algorithms from the developers’ own area of interest can be incorporated into the exercises.
More information: course information (PDF).
Feasibility studies
To help assess whether GPU computing is what your application needs, we offer evaluation and feasibility studies. From quick yet incisive reviews to full in-depth studies, we analyze your challenges and aims and give you the data needed to make the right decisions about further GPU development.
Technologies
GPU computing (formerly GPGPU) uses graphics processors for general computing, giving both substantially higher performance and better power efficiency than CPUs. The two main frameworks are Nvidia’s CUDA and the OpenCL standard, and we work with both. HPC has also developed Xtream, an open-source CUDA C++ framework for developer productivity.