Orange Forum • View topic - GSOC: Support for parallel computation for scripting/backend

GSOC: Support for parallel computation for scripting/backend

General discussions about Orange and with Orange connected things (data mining, machine learning, bioinformatics...).

GSOC: Support for parallel computation for scripting/backend

Postby Jan » Sun Mar 25, 2012 1:33


I am currently a 2nd year master student at Vrije Universiteit Amsterdam and I would very much like to participate in GSOC 2012. Being enrolled in High Performance and Distributed computing master your project proposal "Support for parallel computation for scripting/backend" naturally caught my attention.

However, I do not have any experience with grid computing. My current area of interest is parallel computing on GPUs. Thus, my question is whether you would be interested in parallelization on GPUs or many core CPUs using OpenCL?

Thank you in advance.

Kind regards,


Re: GSOC: Support for parallel computation for scripting/bac

Postby Anze » Sun Mar 25, 2012 20:44

Your idea sound promising. I don't know much about OpenCL so I will not be able to help you on that part of the project. But I can answer your Orange related questions. There are however some things I would like to know, before I say that this is a good idea.

* Orange stores data in its own data format, which is highly compatible with python, but probably nothing else. If Orange would use OpenCL would you need to replace data representation with a new one? If so, how costly would this transformation be?
(Orange 3.0 will store data in a new format, which should be numpy-compatible. If this helps, you proposal can target 3.0 branch of orange).

* How hard would it be to modify existing algorithms to take advantage of this parallelization. Would it require programming in C or would python scripts suffice?

* Are there existing implementations of machine learning algorithms (libraries) that we could take advantage of, if we used OpenCL?


Return to General Discussions