[hpc-announce] Call for Participation: HiPEAC 2018 Conference: rCUDA Tutorial

Federico Silla fsilla at upv.es
Thu Dec 21 07:26:43 CST 2017

[Apologies if you got multiple copies of this email]

HiPEAC 2018 Conference, January 22-24, 2018, Manchester, United Kingdom

Call for Participation for Tutorial:
rCUDA: Greener data centers thanks to remote GPU virtualization


Tuesday, January 23, 2018, from 14:00 to 17:30
Room 5 @ Manchester Central Convention Center

rCUDA is a middleware that allows executing an application accelerated 
with GPUs
in a cluster node that does not own a GPU. Instead, the application will 
make use of
GPUs installed in other nodes of the cluster. In this way, rCUDA allows, 
for instance,
that microservers in a data center execute GPU-accelerated applications 
by using GPUs
installed in regular servers. Additionally, rCUDA allows increasing 
cluster throughput
by 40% and also reducing energy consumption by 40%.

This tutorial presents the benefits of using rCUDA (www.rcuda.net). The 
tutorial also
includes a hands-on part where it is presented how to install and use 
this freely available
middleware. Finally, attendees will be able to exercise with rCUDA by 
connecting to a
real cluster located in Universitat Politecnica de Valencia. This 
cluster includes several
nodes with GPUs as well as the rCUDA software.

More information about the program of the tutorial can be found at:

More information about registration, accommodation and travel
information can be found at the HiPEAC 2018 Conference website:

More information about the hpc-announce mailing list