Services: Helping You Accelerate Towards Success

Acceleware assists customers with their high performance software needs in many different capacities:

  • Software Development and Porting
  • Acceleration Assessment
  • Mentoring


Acceleware migrates your existing code to your target hardware platform.  The reference code can be written in almost any language including C/C++, Java, Fortran, Python, Matlab, and others.  As part of the development process, Acceleware profiles and optimizes the code to ensure that the best possible performance is realized.

Computer Aided Engineering

Computer Aided Engineering

Implemented a GPU based LDLT factorization that improved performance 2-3x over the multi-threaded CPU solution




The acceleration assessment is designed to provide our clients with recommendations which answer the ultimate questions;

  • How fast will it go?
  • How long will the project take?
  • How much will it cost?

For clients who are new to the high performance world the preliminary assessment will also help answer the technical questions; 

  • What hardware platform should I consider?  (Clusters, GPUs, multi-core CPUs, FPGAs, Xeon Phi, AMD vs. NVIDIA)
  • What software development platform is best for me? (OpenCL vs. OpenMP, OpenCL vs. CUDA, MPI)

CT Reconstruction

Accelerated reconstructions from CT scanners by a factor of 30x



For customers who have already ported their code to the desired platform, but would like a second opinion to assess how well the code was written, Acceleware provides mentoring services:

  • On-demand support for your software questions
  • Assistance on specific performance questions relating to ported software
  • Code review and profiling of migrated code
  • Recommendations and feedback on additional optimizations and future development

Option Pricing Models

Improved performance of option pricing by 50x over a single threaded CPU implementation



Acceleware is experienced in working with a wide range of programming languages, platforms and 3rd party tools including:

  • Platforms: Multi-core CPUs, GPUs, Xeon Phi Coprocessors and FPGAs
  • Languages: CUDA, OpenCL, C++, Fortran
  • Multi-threading: MPI, OpenMP, Pthreads, Intel MKL
  • 3rd Party tools: MATLAB, Boost