Eventil - Find Tech Events
Official mobile app
FREE - In Google Play
Official mobile app
FREE - In App Store

Simple Code for High Performance: Utilizing Heterogenous Platforms in F#

Sep 3, 2014 · Zürich, Switzerland

Hi all,

this meetup is focused on high-performance/parallel computing on heterogenous CPU/GPU platforms. Three speakers present original research and demonstrate proven code from successful real-world solutions.

FSCL: Homogeneous Programming and Execution on Heterogeneous Platforms (by Gabriele Cocco, Pisa University)

FSCL is an F#-based framework for OpenCL-compatible platforms. It was created by Gabriele for his current PhD thesis. It aims to balance ease of use, expressiveness, and extensibility, while keeping performance close to native OpenCL C/C++. Gabriele discusses how he has approached the goal with a carefully designed abstract programming layer and runtime support.

Professional GPU Development with F# and the Alea Compiler Suite (by Daniel Egloff, QuantAlea)

Daniel presents the newest version of the high productivity GPU development tool chain Alea based on F# and shows how it can be used to reduce the development time of GPU accelerated applications in .NET. He covers new features such as debugging, profiling, and source code correlation, and gives an outlook on the future development plans to support multiple platforms, mobile GPUs, and other .NET languages.

New Abstractions for Radically Simplified GPU Programming in F# and .NET (by Luc Bläser, HSR, based on joint work with QuantAlea)

Many programmers still leave the massive GPU parallel power unused – be it because of lacking experience in CUDA, OpenCL etc. technologies or because of limited time and budget. The joint project aims to drastically simplify GPU parallelization for the broad group of .NET programmers. By introducing a reactive dataflow programming model, complex computations can be easily and rapidly composed of a set of prefabricated and customizable operations. The underlying F#-based runtime system automatically translates this abstract model to efficient GPU-parallel code.

I would like to thank the three presenters for allocating the time to make this combined event possible.

If you come for the first time, please download this pdf to make sure you find the meetup location inside Zurich Main Station. As always, a fee of CHF 10.-- will be charged to cover part of the expenses.

Enjoy the rest of the summer, and see you soon at an inspiring evening.

Cheers, Marc

Event organizers

Are you organizing Simple Code for High Performance: Utilizing Heterogenous Platforms in F#?

Claim the event and start manage its content.

I am the organizer

based on 0 reviews