Lehr- und Forschungseinheit Informatik X
Rechnertechnik und Rechnerorganisation 

The 1st Workshop on Co-Scheduling of HPC Applications


January 19, 2016

Organised by: Carsten Trinitis and Josef Weidendorfer Technische Universität München, Institut für Informatik

Co-located with HiPEAC 2016 Prague, Czech Republic, January 18-20, 2016

Supported by


The task of a high performance computing system is to carry out its calculations (mainly scientific applications) with maximum performance and energy efficiency.  Up until now, this goal could only be achieved by exclusively assigning an appropriate number of cores/nodes to parallel applications. As a consequence, applications had to be highly optimised in order to achieve even only a fraction of a supercomputer's peak performance which required huge efforts on the programmer side.

This problem is expected to become more serious on future exascale systems with millions of compute cores. Many of today's highly scalable applications will not be able to utilise an exascale system's extreme parallelism due to node specific limitations like e.g. I/O bandwidth. Therefore, to be able to efficiently use future supercomputers, it will be necessary to simultaneously run more than one application on a node. To be able to efficiently perform co-scheduling, applications must not slow down each other, i.e. candidates for co-scheduling could e.g. be a memory-bound and a compute bound application.

Within this context, it might also be necessary to dynamically migrate applications between nodes if e.g. a new application is scheduled to the system.

In order to be able to monitor performance and energy efficiency during operation, additional sensors are required. These need to be correlated to running applications to deliver values for key performance indicators.

Main topics

Exascale architectures, supercomputers, scheduling, performance sensors, energy efficiency, task migration

Workshop Programme

10:00 Workshop opening
by Carsten Trinitis and Josef Weidendorfer
10:00 - 13:00 Session 1
Chair: Carsten Trinitis
10:00 Keynote
New Technologies for High Performance Computing
Christopher Dahnken, Intel Corporation 
10:35 A resource-centric application classification approach
Alexandros-Herodotos Haritatos, Konstantinos Nikas, Georgios Goumas, and Nectarios Koziris
11:00 (coffee break)
11:30 - 11:55
Dynamic Process Management with Allocation-internal Co-Scheduling towards Interactive Supercomputing
Carsten Clauss, Thomas Moschny, and Norbert Eicker
11:55 - 12:20
Detailed Characterization of HPC Applications for Co-Scheduling
Josef Weidendorfer and Jens Breitbart
12:20 - 12:45
Terrible Twins: A Simple Scheme to Avoid Bad Co-Schedules
Andreas de Blanche and Thomas Lundqvist

13:00 - 14:00
Lunch Break

14:00 - 15:30 Session 2
Chair: Josef Weidendorfer
14:00 - 14:25
Implications of Process-Migration in Virtualized Environments
Simon Pickartz, Jens Breitbart, and Stefan Lankes
14:25 - 14:50
Impact of the Scheduling Strategy in Heterogeneous Systems That Provide Co-Scheduling
Tim Süß, Nils Döring, Ramy Gad, Lars Nagel, André Brinkmann, Dustin Feld, Eric Schricker, and Thomas Soddemann

14:50 - 15:25
Panel Discussion
Lame Duck or Silver Bullet? Co-Scheduling for High Performance Computing

15:25 - 15:30

Important Dates

Paper Title and Abstract due: November 1, 2015

Paper Submission deadline: November 15, 2015

Notification: December 5, 2015

Camera ready: December 15, 2015

Workshop: January 19, 2016

Please register for the conference here.

Paper submission

Please submit your paper here.

Carsten Trinitis and Josef Weidendorfer
Lehrstuhl für Rechnertechnik und Rechnerorganisation
Institut für Informatik
Technische Universität München
Boltzmannstr. 3, 85748 Garching bei München
Phone: +49 89 289 18458

E-Mail: {Carsten.Trinitis | Josef.Weidendorfer} <at> tum.de