Universität Paderborn Universität Paderborn
Paderborn Center for Parallel Computing

Go to first section

Paderborn Center for Parallel Computing

The Paderborn Center for Parallel Computing (PC²) is a scientific institute of Paderborn University. Since 1991 it is a competence center for parallel and distributed computing of the state North Rhine-Westfalia (NRW). Its mission is the operation of transregional computing centers and the development of technologies and applications within national and international research projects. The PC² considers itself an interdisciplinary research center, which—together with users from the areas of computer science, mathematics, engineering and natural sciences—works on scientific and industrial issues regarding high-performance computing and datacenter management.

HPC Expertise

  • Provisioning of HPC infrastructure and services for computational sciences, where our main users are coming from our university's profile areas "Optoelectronics and photonics" and "Lightweight design with hybrid systems" (material sciences).

  • Development of new methods for HPC simulation codes and contribution of these codes to widely used open source codes (e.g., CP2K, ls1 MarDyn, ms2, QuantumESPRESSO, ADOL-C).

  • Computing systems research for energy-efficient high-performance computing with an emphasis on heterogeneous and accelerated computing with Field-Programmable Gate Arrays (FPGAs) and Manycore architectures.


HPC Systems

OCuLUS - Clustervision CV-AIRE5

since June 2013 - learn more - Top500.org
GPU-Nodes MPP-System SMP-System Tier 3 - Cluster Xeon-Phi-Nodes
240 TFlop/s
Peak Performance
616 Nodes
QDR-Infiniband
47 TB
Main Memory
9,920 CPU Cores
Intel
32 GPGPUs
Nvidia
8 Many Core Processors
Intel

XCL FPGA-Cluster - Arbor Workstation 2HE C612 INTEL

since May 2017 - Top500.org
FPGA-Nodes
8 Nodes
1 TB
Main Memory
32 CPU Cores
Intel
16 FPGAs
Alpha Data

File Systems

WORK file system

OCuLUS (FhGFS)

HDD Symbol
500 TB
disk storage

Projects

Domain-specific programming and target-aware compiler infrastructure for algorithms on unstructured grids
BMBF
The goal of HighPerMeshes is to develop a practically usable domain-specific framework for the efficient, parallel and scaling implementation of iterative algorithms on unstructured grids.
23%
27 months remaining
Software Parallele Programming
Performance and Efficiency in HPC with Custom Computing
DFG
PerficienCC improves the use of application specific hardware accelerators - FPGA - through specific implementations and analysis.
15%
30 months remaining
Energy Efficiency Performance Analysis

Contact

Prof. Dr. Christian Plessl
Universität Paderborn
Paderborn Center for Parallel Computing
Warburger Str. 100
33098 Paderborn
Germany

 Send a message to Paderborn Center for Parallel Computing (PC²)

Legal Entity

Universität Paderborn

Die Universität Paderborn orientiert sich an der Leitidee der "Universität der Informationsgesellschaft". Die starke Informatik und deren Anwendungsfelder sowie die interdisziplinäre Durchdringung vieler Disziplinen durch informationstechnologische Aspekte bilden für diese Ausrichtung zwar eine hervorragende Grundlage; die Hochschule will aber mehr erreichen: Sie möchte die naturwissenschaftlich-technische Entwicklung der Informationsgesellschaft vorantreiben, sie kritisch begleiten, gleichzeitig den Blick für die beständigen Werte unserer Kultur öffnen, aber auch die sich in der Informations- oder Wissensgesellschaft bietenden Chancen nutzen. Das Lehr- und Forschungsprofil umfasst in den fünf Fakultäten ein breit gestreutes Fächerspektrum in den Geistes-, Wirtschafts-, Natur- und Ingenieurwissenschaften.

Contact

Universität Paderborn
Warburger Str. 100
33098 Paderborn
Germany