High-Performance Computing with GPUs

High-Performance Computing with GPUs

QuantAlea

High productivity GPU development solution and compiler for .NET

QuantAlea, a fully owned subsidiary of InCube, is a Swiss-based software engineering company specialized in high performance and GPU computing technologies.

GPUs are massively parallel processors with thousands of small but highly efficient cores designed for handling multiple tasks simultaneously.

With Alea GPU, you can take advantage of this processing power to accelerate .NET and Mono applications in a simple and efficient way on Windows, Linux and Mac OS X. You develop your GPU code with the .NET language and tools you already know. The Alea GPU runtime system efficiently handles execution on the GPU and all the memory management. 

Alea GPU at a glance

Features: 

  • Extending Microsoft’s modern language F# to first class CUDA language
  • Simple usage from any .NET language
  • Strongly tooled with full debugging, profiling and IntelliSense support
  • Integrated GPU profiling including source code correlation for performance analysis down to individual source code lines
  • Compiler and linker integrated into F# and callable through an API
  • Covering most of the advanced CUDA 6.x features
  • Extensible complier built on top of the modern LLVM compiler technology

Your benefits: 

  • Improve your GPU coding productivity significantly
  • Reduce boiler plate code for data management and data transfer, or GPU resource management
  • Ideal for rapid prototyping of GPU algorithms at all levels of sophistication
  • Experiment with GPU scripting in the F# interactive
  • No compromise on speed: GPU code executes as fast as compiled CUDA C/C++
  • Generate GPU code dynamically at runtime, allowing new application scenarios
  • Simple installation and deployment, single assembly installable from NuGet or MSI package
  • Very light footprint, the complete solution is approximately 20 MB
  • Packaged with many useful algorithms such as random number generators, parallel primitives, linear system solvers and more