StreamComputing

You are here: FrontPage > Training > Self-study > Tools

Tools

July 2011. This page needs a major update!

There is still a lot missing, but I have not included all links I have in my bookmarks. Please contact us, if you have anything to add.

Wrappers

There are wrappers for any language, also to simplify tasks.

• JavaCL. Java/Scala-wrapper, created by Ché zOlive.
• JOCL, created by Michael Bien.
• Big N Computing SOCL.

Libraries

• ViennaCL, a linear algebra library for computations on GPUs written in C++ and based on OpenCL.

Compilers

Alternatives to OpenCL. All nice projects to read about.

• Sh, a predecessor to the RapidMind Multi-core Development Platform, metaprogramming language for programmable GPUs.
• Ocelot, a dynamic compilation framework for heterogeneous systems, accomplishing this by providing various backend targets for CUDA programs. Ocelot currently allows CUDA programs to be executed on NVIDIA GPUs and x86-CPUs at full speed without recompilation.
• Tsunami, a concise parallel-programming language along with a (cross-)compiler for that language that produces the equivalent optimized OpenCL code. This project is still in alpha, but I will blog about it when it has taken more form.
• StreamIt, a programming language and a compilation infrastructure, specifically engineered for modern streaming systems.

Debuggers

• Gremedy gDEBugger.

• Newsletter
  Sign up for our OpenCL Newsletter

  Email Address*

  First Name

  Last Name

  * = required field
• Previous Projects

  Sensor algorithm development By providing more statistics on the output-channels, an algorithm could be developed which gave a sensor-device a good approximation of the results of its big brother multi-sensor.

  
  

  Investment computations caching optimisation - Brought the total processing time of business analytics software from 2 hours to under 30 seconds. A team of 5 people did not have to wait for a long rerun if something went wrong.

  
  

  PDF-checking batch speed-up - By making use of resources more efficiently, the throughput of digital book processing increased by 6 times. The needed amount of processed books per month could be reached using existing hardware.

• Time at the StreamComputing office
  
  
  Amsterdam

  
  
• Direct contact
  
  Want to ask a quick question to Vincent Hindriksen personally via chat? It is completely anonymous! Note that this service only works when the status is or .
  Alternatively you can chat or call via Skype (vhindriksen).
  
  Prefer contact via e-mail? Go to the contact-page and fill in the form. Response within 48 hours!
• Follow us too
  Follow @StreamComputing
• Consultancy & Training
  Need your software to be faster?
  Need training to write GPU-code?
  Don't wait, call Vincent Hindriksen now on how to improve the computation-speed of your algorithms.
  +31 6454 00 456
  Or let us contact you!