Recorded Thursday, February 9, 2012

This webinar is targeted at researchers and companies working on compute-intensive problems and who are interested in harnessing the processing power of hardware accelerators, such as GPUs, with a minimum coding effort. In wireless telecommunications, typical applications that could benefit from this technology are cyclostationary signal analysis, MIMO processing, or network simulations. The webinar explores different high-level programming models for accelerating code on multi-core CPUs and GPUs - directives-based or dataflow-based - and compares the associated development tools. The achievable performance gains are demonstrated using a specific example - a simulation of self-organising wireless network nodes in a dynamic channel allocation scenario.

Presented by: Hicham Lahlou, CTVR

(Webinar) (Slides)