Overview
Topic status: We're looking for students to study this topic.
Project Summary
This project concerns the development of an open source C++ linear algebra library with MATLAB-like syntax. The library is written using C++ templates, allowing for efficient compile-time dispatch to optimised implementations for either CPU or GPU hardware, depending on the data type. The implementations themselves forward to the popular open source GPU library Thrust to do the actual computation.
The core functionality of the library is written and working well. The student will assist in its completion and release, including testing existing functionality, adding new functionality and developing example programs and documentation.
Expected outcomes, applications and/or benefits
The release of version 1 of an open-source C++ linear algebra library with MATLAB-like syntax that performs efficiently on both CPU and GPU architectures.
The student will gain valuable experience in C++ template programming and programming for GPUs; and also in open source software development.
Required student skills/experience
Experience programming in C++, particularly in using the C++ standard library, and in template programming. Experience with MATLAB and/or GPU programming helpful but not essential.
- Study level
- Vacation research experience scholarship
- Supervisors
- QUT
- Organisational unit
Science and Engineering Faculty
- Research area
- Keywords
- open source, linear, algebra, library
- Contact
- Contact the supervisor for more information
Dr Timothy Moroney
