Nathalie Rauschmayr, Optimization of LHCb Applications for Multi- and Manycore Job Submission

→ Europe/Zurich
28/S-029 (CERN)

28/S-029

CERN

35
Show room on map
Description

WhiteArea lectures' twiki HERE

Nowadays, all Computing Grid resources consist of multi- and manycore processors and the main question is how these can be used more efficiently. Therefore, optimization must be applied at many different levels. In the context of cloud computing, experiments have full control over virtual machines and therefore they can arbitrarily configure the operating system such that their jobs profit the most. Another step is the parallelization of software, in order to reduce memory requirements, runtime and to manage concurrent accesses to system resources. Additionaly, scheduling mechanisms must be implemented in order to define the best degree of parallelism for each job. This is necessray because jobs might not scale sufficiently well on large number of cores, which can drastically reduce overall throughput.


Roοm booked by MariaD on 2014/05/07

Slides
The agenda of this meeting is empty