Description:
Dataflow computing is a novel way of performing computation, fundamentally different from instruction-driven computing with conventional CPUsand GPUs. Dataflow computers focus on optimizing the movement of data in an application and exploit massive parallelism among thousands of tiny ‘dataflow cores’ to provide orders of magnitude benefits in performance, rack space and power consumption.

This training will introduce the concepts behind dataflow computing and will present the Multiscale Dataflow High Performance Computing solution based on Maxeler's Dataflow Engines (DFEs) and MaxCompiler. It will showcase how conventional software can be ported and optimized on the DFEs and will feature a hands-on session so that participants can experiment with the technology.

Trainer:
Georgi Gaydadjiev [email protected]
Tel: +44 (0)75 2277 3924

Prerequisites:
Familiarity with Linux, knowledge of programming languages (Java and either of C, Python, R, Matlab)

Learning outcomes:
Participants will understand the main principles behind dataflow computing and will gain hands-on experience with concrete implementations.

https://events.prace-ri.eu/event/499/