Skip to main content

Performance Optimisation and Productivity 2

Deliverables

Plan for targeting SMEs

This deliverable will build on the work done in POP1 and will identify the targeted strategy that will be used in the project to identify and attract SMEs to become POP users.

First Report on Proof-of-Concept

This deliverable will contain a description of the proof-of-concept work during the first half of the project, including a public version of the reports and a summary of the findings and recommendations to the customers. It will also include recommendations for tool developers and programming model standardization bodies that may arise from the studies as well as recommendations from mock-up tests which should be used within the training activities.

Customer Feedback Methodology

Updated customer feedback measurement methodology.

First POP dissemination and training report

This report will summarize the dissemination, cooperation and training activities and events for the first 18 months of the project.

First report on methodology development and tool improvement

Report with two sections: one on the extensions made to the methodology used in the performance assessment, and one on the improvements made to the tools.

Plan for targeting CoEs

This deliverable will set out our strategy for building and maintaining collaborations with the other Centres of Excellence alongside the CSA.

Customer Feedback Measurement

This deliverable will summarize the findings of the Customer Advocate during the first 18 months of the project. It will also include the actual suggestions made to the internal Operational Management meetings of the projects, the actions carried-out to follow-up the POP customers, the Customer Advocate impression of whether and how such suggestions did influence the actual operation.

First Business Development and Sustainability Review

This deliverable will update and review our progress in business development for the first half of the project. Including key events, markets targeted, SME conversion, CoE collaboration and progress on the KPIs. This data will then be used to map out the business development plan for the rest of the project. This deliverable will also review our approach to sustainability in light of the improvements made in the project.

Co-design repository structure

The deliverable will be a prototype of the repository structure.

First Report on Analysis

This deliverable will contain a description and statistics of the cases analysed during the first half of the project, including a public version of the reports and a summary the findings and recommendations to the customers. It will also include recommendations for tool developers that may arise from the studies.

First co-design repository

The deliverable will be the POP repository filled with global data gathered from POP1 and POP2 reports till month 18 including detailed statistics on the relevance of the different efficiency factors in the POP methodology. It will also include a first set of 8 microbenchmarks characterizing behaviours found in real applications.

Data Management Plan (DMP)

Analysis of the main elements of the data management policy with regard to all the datasets generated by the project. The DMP will include a table specifying how the data will be exploited, shared for verification and re- use. In case the table cannot be provided a justification will be provided.

Publications

Collectives in hybrid MPI+MPI code: Design, practice and performance

Author(s): Huan Zhou, José Gracia, Naweiluo Zhou, Ralf Schneider
Published in: Parallel Computing, 99, 2020, Page(s) 102669, ISSN 0167-8191
Publisher: Elsevier BV
DOI: 10.1016/j.parco.2020.102669

Towards blood flow in the virtual human: efficient self-coupling of HemeLB

Author(s): J. W. S. McCullough; Robin A. Richardson; Alex Patronis; Alex Patronis; Rene Halver; R. Marshall; Martin Ruefenacht; Brian J. N. Wylie; Thomas Odaker; Markus Wiedemann; Bryn A. Lloyd; Esra Neufeld; Godehard Sutmann; Godehard Sutmann; Anthony Skjellum; Dieter Kranzlmüller; Peter V. Coveney; Peter V. Coveney
Published in: Interface focus , 11 (1) , Article 20190119. (2021), 3, 2020, ISSN 2042-8901
Publisher: Royal Society
DOI: 10.3929/ethz-b-000465832

A Generic Performance Analysis Technique Applied to Different CFD Methods for HPC

Author(s): Marta Garcia-Gasulla, Fabio Banchelli, Kilian Peiro, Guillem Ramirez-Gargallo, Guillaume Houzeaux, Ismaïl Ben Hassan Saïdi, Christian Tenaud, Ivan Spisso, Filippo Mantovani
Published in: International Journal of Computational Fluid Dynamics, 34/7-8, 2020, Page(s) 508-528, ISSN 1061-8562
Publisher: Taylor & Francis
DOI: 10.1080/10618562.2020.1778168

Tools for GPU Computing - Debugging and Performance Analysis of Heterogenous HPC Applications

Author(s): Michael Kobloch, Bernd Mohr
Published in: Supercomputing Frontiers and Innovations, 7/1, 2020, Page(s) 91-111, ISSN 2313-8734
Publisher: South Ural State University
DOI: 10.14529/jsfi200105

Performance and energy consumption of HPC workloads on a cluster based on Arm ThunderX2 CPU

Author(s): Filippo Mantovani, Marta Garcia-Gasulla, José Gracia, Esteban Stafford, Fabio Banchelli, Marc Josep-Fabrego, Joel Criado-Ledesma, Mathias Nachtmann
Published in: Future Generation Computer Systems, 112, 2020, Page(s) 800-818, ISSN 0167-739X
Publisher: Elsevier BV
DOI: 10.1016/j.future.2020.06.033

Dynamic Runtime and Energy Optimization for Power-Capped HPC Applications

Author(s): Bo Wang, Christian Terboven, Matthias Müller
Published in: Advances in Parallel Computing, Volume 36: Parallel Computing: Technology Trends, 2020, Page(s) 441 - 452
Publisher: IOS Press
DOI: 10.3233/apc200070

Using performance analysis tools for parallel-in-time integrators -- Does my time-parallel code do what I think it does?

Author(s): Speck, Robert; Knobloch, Michael; Lührs, Sebastian; Gocht, Andreas
Published in: Parallel-in-Time Integration Methods. PinT 2020. Springer Proceedings in Mathematics & Statistics, 1, 2021, Page(s) 51-80, ISBN 978-3-030-75933-9
Publisher: Springer
DOI: 10.1007/978-3-030-75933-9_3

Performance Prediction for Power-Capped Applications based on Machine Learning Algorithms

Author(s): Bo Wang, Jannis Klinkenberg, Daniel Ellsworth, Christian Terboven, Matthias Muller
Published in: 2019 International Conference on High Performance Computing & Simulation (HPCS), 2019, Page(s) 842-849, ISBN 978-1-7281-4484-9
Publisher: IEEE
DOI: 10.1109/hpcs48598.2019.9188144

Analyzing the Efficiency of Hybrid Codes

Author(s): Judit Gimenez, Estanislao Mercadal, German Llort, Sandra Mendez
Published in: 2020 19th International Symposium on Parallel and Distributed Computing (ISPDC), 2020, Page(s) 29-36, ISBN 978-1-7281-8946-8
Publisher: IEEE
DOI: 10.1109/ispdc51135.2020.00014

Exascale potholes for HPC: Execution performance and variability analysis of the flagship application code HemeLB

Author(s): Brian J. N. Wylie
Published in: 2020 IEEE/ACM International Workshop on HPC User Support Tools (HUST) and Workshop on Programming and Performance Visualization Tools (ProTools), 2020, Page(s) 59-70, ISBN 978-1-6654-2280-2
Publisher: IEEE Xplore
DOI: 10.1109/hustprotools51951.2020.00014

MPI Detach - Asynchronous Local Completion

Author(s): Joachim Protze, Marc-André Hermanns, Ali Demiralp, Matthias S. Müller, Torsten Kuhlen
Published in: 27th European MPI Users' Group Meeting, 2020, Page(s) 71-80, ISBN 9781450388801
Publisher: ACM
DOI: 10.1145/3416315.3416323

Benchmarking of state-of-the-art HPC Clusters with a Production CFD Code

Author(s): Fabio Banchelli, Marta Garcia-Gasulla, Guillaume Houzeaux, Filippo Mantovani
Published in: Proceedings of the Platform for Advanced Scientific Computing Conference, 2020, Page(s) 1-11, ISBN 9781450379939
Publisher: ACM
DOI: 10.1145/3394277.3401847

MPI Collectives for Multi-core Clusters - Optimized Performance of the Hybrid MPI+MPI Parallel Codes

Author(s): Huan Zhou, José Gracia, Ralf Schneider
Published in: Proceedings of the 48th International Conference on Parallel Processing: Workshops, yearly, 2019, Page(s) 1-10, ISBN 9781-450371964
Publisher: ACM
DOI: 10.1145/3339186.3339199

Performance study of HPC applications on an Arm-based cluster using a generic efficiency model

Author(s): Fabio Banchelli, Kilian Peiro, Andrea Querol, Guillem Ramirez-Gargallo, Guillem Ramirez-Miranda, Joan Vinyals, Pablo Vizcaino, Marta Garcia-Gasulla, Filippo Mantovani
Published in: 2020 28th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP), yearly, 2020, Page(s) 167-174, ISBN 978-1-7281-6582-0
Publisher: IEEE
DOI: 10.1109/pdp50117.2020.00032

Cluster of emerging technology: evaluation of a production HPC system based on A64FX

Author(s): Fabio Banchelli, Kilian Peiro, Guillem Ramirez-Gargallo, Joan Vinyals, David Vicente, Marta Garcia-Gasulla, Filippo Mantovani
Published in: 2021 IEEE International Conference on Cluster Computing (CLUSTER), 2021, Page(s) 741-750, ISBN 978-1-7281-9666-4
Publisher: IEEE
DOI: 10.1109/cluster48925.2021.00110

Towards compiler-aided correctness checking of adjoint MPI applications

Author(s): Alexander Huck, Joachim Protze, Jan-Patrick Lehr, Christian Terboven, Christian Bischof, Matthias S. Muller
Published in: 2020 IEEE/ACM 4th International Workshop on Software Correctness for HPC Applications (Correctness), 2020, Page(s) 40-48, ISBN 978-0-7381-1044-8
Publisher: IEEE
DOI: 10.1109/correctness51934.2020.00010

TALP - A Lightweight Tool to Unveil Parallel Efficiency of Large-scale Executions

Author(s): Victor Lopez, Guillem Ramirez Miranda, Marta Garcia-Gasulla
Published in: Proceedings of the 2021 on Performance EngineeRing, Modelling, Analysis, and VisualizatiOn STrategy, 2021, Page(s) 3-10, ISBN 9781450383875
Publisher: ACM
DOI: 10.1145/3452412.3462753

A Case Study on Addressing Complex Load Imbalance in OpenMP

Author(s): Fabian Orland, Christian Terboven
Published in: OpenMP: Portable Multi-Level Parallelism on Modern Systems - 16th International Workshop on OpenMP, IWOMP 2020, Austin, TX, USA, September 22–24, 2020, Proceedings, 12295, 2020, Page(s) 130-145, ISBN 978-3-030-58143-5
Publisher: Springer International Publishing
DOI: 10.1007/978-3-030-58144-2_9

Task Inefficiency Patterns for a Wave Equation Solver

Author(s): Holger Schulz, Gonzalo Brito Gadeschi, Oleksandr Rudyy, Tobias Weinzierl
Published in: OpenMP: Enabling Massive Node-Level Parallelism - 17th International Workshop on OpenMP, IWOMP 2021, Bristol, UK, September 14–16, 2021, Proceedings, 12870, 2021, Page(s) 111-124, ISBN 978-3-030-85261-0
Publisher: Springer International Publishing
DOI: 10.1007/978-3-030-85262-7_8

A Picture Is Worth a Thousand Numbers—Enhancing Cube’s Analysis Capabilities with Plugins

Author(s): Michael Knobloch, Pavel Saviankou, Marc Schlütter, Anke Visser, Bernd Mohr
Published in: Tools for High Performance Computing 2018 / 2019 - Proceedings of the 12th and of the 13th International Workshop on Parallel Tools for High Performance Computing, Stuttgart, Germany, September 2018, and Dresden, Germany, September 2019, 2021, Page(s) 237-259, ISBN 978-3-030-66056-7
Publisher: Springer International Publishing
DOI: 10.1007/978-3-030-66057-4_13

Score-P and OMPT: Navigating the Perils of Callback-Driven Parallel Runtime Introspection

Author(s): Christian Feld, Simon Convent, Marc-André Hermanns, Joachim Protze, Markus Geimer, Bernd Mohr
Published in: OpenMP: Conquering the Full Hardware Spectrum - 15th International Workshop on OpenMP, IWOMP 2019, Auckland, New Zealand, September 11–13, 2019, Proceedings, 11718, 2019, Page(s) 21-35, ISBN 978-3-030-28595-1
Publisher: Springer International Publishing
DOI: 10.1007/978-3-030-28596-8_2

A Study of Memory Anomalies in OpenMP Applications

Author(s): Lechen Yu, Joachim Protze, Oscar Hernandez, Vivek Sarkar
Published in: OpenMP: Portable Multi-Level Parallelism on Modern Systems - 16th International Workshop on OpenMP, IWOMP 2020, Austin, TX, USA, September 22–24, 2020, Proceedings, 12295, 2020, Page(s) 328-342, ISBN 978-3-030-58143-5
Publisher: Springer International Publishing
DOI: 10.1007/978-3-030-58144-2_21

Datasets