Planned intervention: On Wednesday April 3rd 05:30 UTC Zenodo will be unavailable for up to 2-10 minutes to perform a storage cluster upgrade.
Published September 15, 2016 | Version v1
Report Open

Performance studies on different accelerators using OpenCL

  • 1. CERN openlab Summer Student
  • 2. Summer Student Supervisor

Description

Project Specification

The High Luminosity LHC (HL-LHC) is a project to increase the luminosity of the Large Hadron Collider to 5*1034 cm-2 s-1. The CMS experiment is planning a major upgrade in order to cope with an expected average number of overlapping collisions per bunch crossing of 140. The dataset sizes will increase by several orders of magnitude and so will be the request for larger computing infrastructure. The complete exploitation of a machine capability is desirable, if not a requirement, that should anticipate a request for new hardware resources to the funding agencies. Furthermore, energy consumption for computing is becoming more and more an important voice into European data center’s budget. The exploitation of Intel integrated accelerators like graphics processors or FPGAs, which are and will be part of our machines, will allow us to achieve much higher energy efficiency and higher performance. Furthermore, MIMD architectures like Kalray’s Massively Parallel Processor Array could prove useful as embedded solutions in real-time environments like the experiment trigger farms. All these accelerators can be programmed using OpenCL. The aim of the project is to study the performance and the power efficiency of these accelerators when executing some kernels which are part of the reconstruction of CMS experiment events using the CMS software framework.

 

Abstract

This report introduces the OpenCL API and programming language and describes implementations using OpenCL of several kernels used for particle track reconstruction in the CMS software framework. The first part are kernels for construction and search in the context of a k-d tree data structure. The second part is a set of kernels for building possible tracks out of pairs of hits in the silicon tracker. Several OpenCL platforms are tested and benchmarked.

Files

report_dernst.pdf

Files (750.8 kB)

Name Size Download all
md5:f967df0beab45b90dc6a09ffbcff1415
750.8 kB Preview Download