Publications and presentations

Journal publications and conference proceedings

• Exploiting Nested Task-Parallelism in the H-LU Factorization, Rocío Carratalá-Sáez (UJI), Sven Christophersen, José I. Aliaga (UJI), Vicenç Beltran (BSC), Steffen Börm, Enrique S. Quintana-Ortı́ (UJI), accepted for publication inJournal of Computational Science - PDSEC2017 Special Issue (to appear).
• Interoperability Strategies for GASPI and MPI in Large Scale Scientific Applications, Christian Simmendinger, Dana Akhmetova, Luis Cebamanos, Tiberiu Rotaru, Mirko Rahn, Stefano Markidis, Erwin Laure, Valeria Bartsch, Roman Iakymchuk, International Journal of High Performance Computing Applications, published on 2018-11-14; DOI: 10.1177/1094342018808359 and Zenodo pre-print: 10.5281/zenodo.1206292.
• Driving Asynchronous Distributed Tasks with Events, Nick Brown, Oliver Brown, Mark Bull, Proceedings of 4th Workshop for Open Source Supercomputing (OpenSuCo), SC18, (Dallas, USA, 12 November 2018); https://sc18.supercomputing.org/proceedings/workshops/workshop_files/ws_opensu102s1-file1.pdf.
• Efficient Algorithms for Collective Operations with Notified Communication in Shared Windows, Muhammed Abdullah Al Ahad, Christian Simmendinger, Roman Iakymchuk, Tiberiu Rotaru, Erwin Laure, Stefano Markidis, Parallel Applications Workshop, Alternatives to MPI (PAW-ATM), SC18, (Dallas, USA, 16 November 2018), Zenodo pre-print: DOI: 10.5281/zenodo.1346121.
• Planning for Performance: Enhancing Achievable Performance for MPI through Persistent Collective Operations, Daniel J. Holmes, Bradley Morgan, Anthony Skjellum, Purushotham V. Bangalore, Srinivas Sridharan, Parallel Computing, Vol. 81, Jan 2019, pp. 32-57; DOI: 10.1016/j.parco.2018.08.001.
• Improving the Interoperability between MPI and Task-Based Programming Models, Jesus Labarta, Vicenç Beltran, Antonio J. Peña, Josep M. Perez, Xavier Teruel, Jorge Bellón, Daniel Holmes, Pau Farré and Kevin Sala,  Proceedings of the 25th European MPI Users' Group Meeting, EuroMPI 2018 (Barcelona, Spain, 23-26 September 2018); DOI: 10.1145/3236367.3236382.
• Communication in task-parallel ILU-preconditioned CG solvers using MPI+OmpSs, José I. Aliaga, Maria Barreda, Goran Flegar, Matthias Bollhöfer, Enrique S. Quintana-Ortí, Concurrency and Computation: Practice and Experience, Vol.29(21), pp. 29 - 43, November 2017; DOI: 10.1002/cpe.4280.
• Planning for Performance: Persistent Collective Operations for MPI, Bradley Morgan, Daniel J. Holmes, Anthony Skjellum, Purushotham Bangalore and Srinivas Sridharan, EuroMPI/USA '17 Proceedings of the 24th European MPI Users' Group Meeting (Chicago, USA, 25-28 September 2017), Article no. 4; DOI: 10.1145/3127024.3127028.
• Interoperability of GASPI and MPI in Large Scale Scientific Applications (click here for pre-print version), Dana Akhmetova, Luis Cebamanos, Roman Iakymchuk, Tiberiu Rotaru, Mirko Rahn, Stefano Markidis, Erwin Laure, Valeria Bartsch, Christian Simmendinger, PPAM 2017 - the 12th International Conference on Parallel Processing and Applied Mathematics (Lublin, Poland, 10-13 September 2017), Lecture Notes in Computer Science Vol.10778, pp. 277-287; DOI: 10.1007/978-3-319-78054-2_26.
• PLASMA: Parallel Linear Algebra Software for Multicore Using OpenMP, Jack Dongarra, Mark Gates, Azzam Haidar, Jakub Kurzak, Piotr Luszczek, Panruo Wu, Ichitaro Yamazaki, Asim Yarkhan, Maksims Abalenkovs, Negin Bagherpour, Sven Hammarling, Jakub Šístek, David Stevens, Mawussi Zounon, Samuel Relton, accepted in July 2018 for publication in ACM Transactions on Mathematical Software, 0098-3500; https://www.research.manchester.ac.uk/portal/en/publications/plasma-parallel-linear-algebra-software-for-multicore-using-openmp(b7ac0bfc-54c1-4f3f-abb1-2dfadb75522e).html
• Task-parallel LU factorization of hierarchical matrices using OmpSs, José I. Aliaga, Rocío Carratalá, Ronald Kriemann, Enrique S. Quintana-Ortí, 18th IEEE International Workshop on Parallel and Distributed Scientific and Engineering Computing (PDSEC) 2017 at IPDPS - IEEE International Distributed and Parallel Processing Symposium, (Orlando (FL), USA, 2 June 2017); DOI: 10.1109/IPDPSW.2017.124.
• Performance study of multithreaded MPI and OpenMP tasking in a large scientific code, Dana Akhmetova, Roman Iakymchuk, Orjan Ekeberg, Erwin Laure, High Level Programming Models and Supporting Environments Workshop (HIPS 2017) at IPDPS - IEEE International Distributed and Parallel Processing Symposium, (Orlando (FL), USA, 29 May 2017), published in Proceedings of Parallel and Distributed Processing Symposium Workshops (IPDPSW), 2017, IEEE International, pages 756-765; https://ieeexplore.ieee.org/document/7965119; also preprint version available here.
• MPI Sessions: Leveraging Runtime Infrastructure to Increase Scalability of Applications at Exascale, Daniel Holmes, Kathryn Mohror, Ryan E. Grant, Anthony Skjellum, Martin Schulz, Wesley Bland, Jeffery M. Squyres, Proceedings of the 23rd European MPI Users' Group Meeting, EuroMPI 2016 (Edinburgh, UK, 25-28 September 2016). Slides available at http://www.eurompi2016.ed.ac.uk/sites/default/files/attachments/Holmes-MPI_Sessions.pdf; DOI: 10.1145/2966884.2966915.
• Exploiting Task-Parallelism in Message-Passing Sparse Linear System Solvers Using OmpSs, José I. Aliaga, María Barreda, Matthias Bollhöfer, Enrique S. Quintana-Ortí - paper presented at Euro-Par 2016 (Grenoble, France, 24-26 August 2016) and published in Euro-Par 2016: Parallel Processing - Volume 9833 of the series Lecture Notes in Computer Science pp 631-643; DOI: 10.1007/978-3-319-43659-3_46.

Poster presentations

• INTERTWinE: Progamming Model INTERoperability ToWards Exascale, Roman Iakymchuk - poster presented at the Café Européen de la Recherche at the Teratec Forum 2018 (Palaiseau, France, 19-20 June 2018)
• INTERTWinE: Progamming Model INTERoperability ToWards Exascale, Jakub Sistek - poster presented at HiPEAC18 (Manchester, UK, 22-24 January 2018)
• INTERTWinE: Programming Model INTERoperability ToWards Exascale, Valeria Bartsch - poster presented at the Café Européen de la Recherche at the Teratec Forum 2017 (Palaiseau, France, 27-28 June 2017)
• INTERTWinE: Resource Manager APIs, Xavier Teruel, Vicenç Beltran, Mark Bull, Olivier Aumage, Enrique S. Quintana - research poster presented at ISC 2017 (Frankfurt, Germany, 18-22 June 2017);
• MPI RMA as a directory/cache interoperability layer, Nick Brown, Tiberiu Rotaru, Bernd Lörwald, Vicenç Beltran, Xavier Teruel, Olivier Aumage - research poster presented at ISC 2017 (Frankfurt, Germany, 18-22 June 2017);
• Directory/Cache API for Sharing Data in Distributed Memory Systems, Tiberiu Rotaru, Bernd Lörwald, Mirko Rahn, Nick Brown, Vicenç Beltran Querol, Olivier Aumage, Xavier Teruel - research poster presented at ISC 2017 (Frankfurt, Germany, 18-22 June 2017);
• INTERTWinE: Programming Model INTERoperability ToWards Exascale, Valeria Bartsch - project poster presented at ISC 2017 (Frankfurt, Germany, 18-22 June 2017);
• Leveraging MPI RMA for an efficient directory/cache transport layer, Nick Brown, Tiberiu Rotaru, Bernd Lörwald, Mirko Rahn, Vicenç Beltran, Xavier Teruel, Olivier Aumage - poster and associated lightning talk presented at EuroMPI 2016 (Edinburgh, UK, 25-28 September 2016).

Other presentations

• StarPU and KStar, Synergies between a Task-Based Runtime System and an OpenMP Compiler, Olivier Aumage - invited talk at SIAM Parallel Processing for Scientific Computing (SIAM PP) 2018 Tokyo, Japan, 7–10 March, 2018
• OpenMP Tasks for QR Factorization in PLASMA, Jakub Šístek, Jakub Kurzak, Julien Langou - talk presented at SIAM Parallel Processing for Scientific Computing (SIAM PP) 2018 Tokyo, Japan, 7–10 March, 2018 (http://meetings.siam.org/sess/dsp_talk.cfm?p=89712).
• Using OpenMP tasks for QR factorization in the PLASMA library, Jakub Śístek - talk presented at High Performance Computing in Science and Engineering (HPCSE) 2017, Ostrava, Czech Republic, 22-26 May 2017.

White Paper on 2016 Exascale Applications Workshop

In October 2016, INTERTWinE led the organisation of a two-day Exascale Applications Workshop. This workshop brought together 12 application owners and 4 EC-funded projects (INTERTWinE, EPiGRAM, EXA2CT and AllScale) to discuss application needs in the exascale era. The applications cover a lot of well-established HPC codes as well as upcoming codes: particles-in-cell code for space plasma simulation; CFD simulation for aeronautics or multiphysics; simulation of complex fluids; dense and sparse linear algebra; machine learning; and quantum chemistry.

For a full report on the findings of the workshop, please read our White Paper on the Exascale Applications Workshop.

Second Exascale Applications Workshop (Edinburgh, 2018) - Final Report

In April 2018, INTERTWinE organised the 2nd Exascale Applications Workshop at the University of Edinburgh. The workshop programme was assembled around four consecutive sessions: Asynchronous execution, Interoperability, Usage of libraries, and Distributed tasks.

For more information, please read our Final Report on the workshop.