SAGE

Leistungen

Initial Dissemination and Exploitation Plan
First Review Dissemination, Collaboration and Exploitation Report and Future Plans
State of the art of I/O with MPI and gap analysis

Report analyzing the suitability of the Message Passing programming model for data centric computing and Percipient Storage. It describes the potential of MPI for data centric computing and includes an analysis of MPI limitations that might prevent MPI from achieving Exascale/Exabyte performance. This study will investigate the suitability of new MPI concepts in select SAGE applications.

Global address space concepts for Percipient Storage

Describes the suitability of the PGAS concepts, such as global address space, and remote direct access to diverse memory spaces, to the SAGE architecture. This report also discusses advantages and limitations of this approach.

Final Review Dissemination, Collaboration and Exploitation Report
Object Storage Integrity Checking : Concept and Architecture Report
Analysis framework Requirements specification

Report on the requirements on an analysis framework to deal with data in Exascale. By studying of the SAGE and Apache Flink interfaces, it identifies the parts of the analysis framework that have to be adjusted to exploit a multi-layered storage system.

Analysis framework (Sub-) System Architecture

This deliverable specifies the architecture of the system components that have to change to work with SAGE. It describes the design of these sub-components to use SAGE for data access and output as well as integration with operators.

Report on implementation of SAGE visualisation support

Report that documents the final implementation of the virtual memory hierarchy for visualisation and reports on related experiences with the SAGE architecture.

Performance Profiling & Debugging Tools for SAGE: Final Validation report

Interim Review Dissemination, Collaboration and Exploitation Report and Future Plans
Performance Profiling & Debugging Tools for SAGE: Concept and Architecture Report

Performance Profiling & Debugging Tools for SAGE: Concept and Architecture Report

Validation Report: Message passing & global address space for Data Centric Computing

Report that compares selected programming models for data centric computing and Percipient Storage for the SAGE architecture. It compares the different approaches and investigates a hybrid solution that includes different features from various programming models. The assessment of different programming model effectiveness will be based on evaluation results of selected SAGE applications.

Analysis framework Interface specification

This deliverable specifies the interfaces for Apache Flink for interaction with SAGE. For file input and output, a sequential file access interface will be exposed. Based on availability, an interface for random-access and filter pushdown is evaluated.

Report on Percipient Storage run-time system architecture

Report providing a specification of the run-time system and documenting the requirements for the underlying software and hardware layers.

Project Website
Initial Project Dissemination Material

Veröffentlichungen

SAGE: Percipient Storage for Exascale Data Centric Computing

Autoren: Sai Narasimhamurthy, Nikita Danilov, Sining Wu, Ganesan Umanesan, Stefano Markidis, Sergio Rivas-Gomez, Ivy Bo Peng, Erwin Laure, Dirk Pleiter, Shaun de Witt
Veröffentlicht in: Parallel Computing, 2018, ISSN 0167-8191
Herausgeber: Elsevier BV
DOI: 10.1016/j.parco.2018.03.002

HACC

Autoren: Salman Habib, Joe Insley, David Daniel, Patricia Fasel, Zarija Lukić, Vitali Morozov, Nicholas Frontiere, Hal Finkel, Adrian Pope, Katrin Heitmann, Kalyan Kumaran, Venkatram Vishwanath, Tom Peterka
Veröffentlicht in: Communications of the ACM, Ausgabe 60/1, 2016, Seite(n) 97-104, ISSN 0001-0782
Herausgeber: Association for Computing Machinary, Inc.
DOI: 10.1145/3015569

RTHMS: a tool for data placement on hybrid memory system

Autoren: Ivy Bo Peng, Roberto Gioiosa, Gokcen Kestor, Pietro Cicotti, Erwin Laure, Stefano Markidis
Veröffentlicht in: ACM SIGPLAN Notices, Ausgabe 52/9, 2017, Seite(n) 82-91, ISSN 0362-1340
Herausgeber: Special Interest Group on Computer Graphics, Association for Computing Machinery
DOI: 10.1145/3156685.3092273

Use of massively parallel computing to improve modelling accuracy within the nuclear sector

Autoren: L M Evans, J D Arregui-Mena, P M Mummery, R Akers, E Surrey, A Shterenlikht, M Broggi, L Margetts
Veröffentlicht in: The International Journal of Multiphysics, Ausgabe 10/2, 2016, ISSN 1750-9548
Herausgeber: Multi-Science Publishing Co. Ltd
DOI: 10.21152/1750-9548.10.2.215

Soft real-time analysis of ITER magnetics streaming data using SPECTRE

Autoren: Jonathan Hollocombe, Shaun de Witt, Ivan Lupelli, David Muir, Rob Akers
Veröffentlicht in: Fusion Engineering and Design, Ausgabe 123, 2017, Seite(n) 869-872, ISSN 0920-3796
Herausgeber: Elsevier BV
DOI: 10.1016/j.fusengdes.2017.03.113

Exascale computing and big data

Autoren: Daniel A. Reed, Jack Dongarra
Veröffentlicht in: Communications of the ACM, Ausgabe 58/7, 2015, Seite(n) 56-68, ISSN 0001-0782
Herausgeber: Association for Computing Machinary, Inc.
DOI: 10.1145/2699414

Generating custom code for efficient query execution on heterogeneous processors

Autoren: Sebastian Breß, Bastian Köcher, Henning Funke, Steffen Zeuch, Tilmann Rabl, Volker Markl
Veröffentlicht in: The VLDB Journal, 2018, ISSN 1066-8888
Herausgeber: Springer Verlag
DOI: 10.1007/s00778-018-0512-y

Characterizing the performance benefit of hybrid memory system for HPC applications

Autoren: Ivy Bo Peng, Roberto Gioiosa, Gokcen Kestor, Jeffrey S. Vetter, Pietro Cicotti, Erwin Laure, Stefano Markidis
Veröffentlicht in: Parallel Computing, Ausgabe 76, 2018, Seite(n) 57-69, ISSN 0167-8191
Herausgeber: Elsevier BV
DOI: 10.1016/j.parco.2018.04.007

A Performance Characterization of Streaming Computing on Supercomputers

Autoren: Stefano Markidis, Ivy Bo Peng, Roman Iakymchuk, Erwin Laure, Gokcen Kestor, Roberto Gioiosa
Veröffentlicht in: Procedia Computer Science, Ausgabe 80, 2016, Seite(n) 98-107, ISSN 1877-0509
Herausgeber: ICCS2016
DOI: 10.1016/j.procs.2016.05.301

Preparing HPC Applications for the Exascale Era: A Decoupling Strategy

Autoren: Ivy Bo Peng, Roberto Gioiosa, Gokcen Kestor, Erwin Laure, Stefano Markidis
Veröffentlicht in: 2017 46th International Conference on Parallel Processing (ICPP), 2017, Seite(n) 1-10, ISBN 978-1-5386-1042-8
Herausgeber: IEEE
DOI: 10.1109/ICPP.2017.9

On a Storage System Software Stack for Extreme Scale Data Centric Computing

Autoren: Sai Narasimhamurthy
Veröffentlicht in: 2017 46th International Conference on Parallel Processing Workshops (ICPPW), 2017, Seite(n) 228-230, ISBN 978-1-5386-1044-2
Herausgeber: IEEE
DOI: 10.1109/ICPPW.2017.41

MPI windows on storage for HPC applications

Autoren: Sergio Rivas-Gomez, Roberto Gioiosa, Ivy Bo Peng, Gokcen Kestor, Sai Narasimhamurthy, Erwin Laure, Stefano Markidis
Veröffentlicht in: Proceedings of the 24th European MPI Users' Group Meeting on - EuroMPI '17, 2017, Seite(n) 1-11, ISBN 9781-450348492
Herausgeber: ACM Press
DOI: 10.1145/3127024.3127034

Pipelined Query Processing in Coprocessor Environments

Autoren: Henning Funke, Sebastian Breß, Stefan Noll, Volker Markl, Jens Teubner
Veröffentlicht in: Proceedings of the 2018 International Conference on Management of Data - SIGMOD '18, 2018, Seite(n) 1603-1618, ISBN 9781-450347037
Herausgeber: ACM Press
DOI: 10.1145/3183713.3183734

Optimized on-demand data streaming from sensor nodes

Autoren: Jonas Traub, Sebastian Breß, Tilmann Rabl, Asterios Katsifodimos, Volker Markl
Veröffentlicht in: Proceedings of the 2017 Symposium on Cloud Computing - SoCC '17, 2017, Seite(n) 586-597, ISBN 9781-450350280
Herausgeber: ACM Press
DOI: 10.1145/3127479.3131621

Efficient k-means on GPUs

Autoren: Clemens Lutz, Sebastian Breß, Tilmann Rabl, Steffen Zeuch, Volker Markl
Veröffentlicht in: Proceedings of the 14th International Workshop on Data Management on New Hardware - DAMON '18, 2018, Seite(n) 1-3, ISBN 9781-450358538
Herausgeber: ACM Press
DOI: 10.1145/3211922.3211925

Exploring the Vision Processing Unit as Co-Processor for Inference

Autoren: Sergio Rivas-Gomez, Antonio J. Pena, David Moloney, Erwin Laure, Stefano Markidis
Veröffentlicht in: 2018 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), 2018, Seite(n) 589-598, ISBN 978-1-5386-5555-9
Herausgeber: IEEE
DOI: 10.1109/IPDPSW.2018.00098

NVIDIA Tensor Core Programmability, Performance & Precision

Autoren: Stefano Markidis, Steven Wei Der Chien, Erwin Laure, Ivy Bo Peng, Jeffrey S. Vetter
Veröffentlicht in: 2018 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), 2018, Seite(n) 522-531, ISBN 978-1-5386-5555-9
Herausgeber: IEEE
DOI: 10.1109/IPDPSW.2018.00091

Estimating Join Selectivities using Bandwidth-Optimized Kernel Density Models

Autoren: Martin Kiefer, Max Heimel, Sebastian Bress, Volker Markl
Veröffentlicht in: Proceedings of the 43rd International Conference on Very Large Data Bases, Munich, Germany, 2017, ISSN 2150-8097
Herausgeber: VLDB

Mero: Co-Designing an Object Store for Extreme Scale

Autoren: Nikita Danilov, Nathan Rutman, Sai Narasimhamurthy and John Bent
Veröffentlicht in: 2016
Herausgeber: PDSW

Efficient Storage and Analysis of Genome Data in Databases

Autoren: Sebastian Dorok, Sebastian Breß, Jens Teubner, Horstfried Läpple, Gunter Saake, Volker Markl
Veröffentlicht in: 17th Conference on Database Systems for Business, Technology, and Web (BTW), 2017
Herausgeber: 17th Conference on Database Systems for Business, Technology, and Web (BTW)

Exploring Application Performance on Emerging Hybrid-Memory Supercomputers

Autoren: Peng, I. B., Markidis, S., Laure, E., Kestor, G., & Gioiosa, R
Veröffentlicht in: HPCC 2016, 2016
Herausgeber: HPCC 2016

Soft real-time analysis of ITER magnetics streaming data using SPECTRE

Autoren: Jonathan Hollocombe, Shaun de Witt, Rob Akers
Veröffentlicht in: Fusion Engineering and Design, 2016
Herausgeber: 29th Symposium on Fusion Technology

Exploring Application Performance on Emerging Hybrid-Memory Supercomputers

Autoren: Ivy Bo Peng, Stefano Markidis, Erwin Laure, Gokcen Kestor, Roberto Gioiosa
Veröffentlicht in: 2016 IEEE 18th International Conference on High Performance Computing and Communications; IEEE 14th International Conference on Smart City; IEEE 2nd International Conference on Data Science and Systems (HPCC/SmartCity/DSS), 2016, Seite(n) 473-480, ISBN 978-1-5090-4297-5
Herausgeber: IEEE
DOI: 10.1109/HPCC-SmartCity-DSS.2016.0074

Exploring the Performance Benefit of Hybrid Memory System on HPC Environments

Autoren: Peng, I. B., Markidis, S., Kestor, G., Cicotti, P., Laure, E., & Gioiosa, R.
Veröffentlicht in: ASHES workshop at IPDPS 2017, 2017
Herausgeber: IPDPS

Idle Period Propagation in Message-Passing Applications

Autoren: Ivy Bo Peng, Stefano Markidis, Erwin Laure, Gokcen Kestor, Roberto Gioiosa
Veröffentlicht in: 2016 IEEE 18th International Conference on High Performance Computing and Communications; IEEE 14th International Conference on Smart City; IEEE 2nd International Conference on Data Science and Systems (HPCC/SmartCity/DSS), 2016, Seite(n) 937-944, ISBN 978-1-5090-4297-5
Herausgeber: IEEE
DOI: 10.1109/HPCC-SmartCity-DSS.2016.0134

Preemptive Data Distribution Infrastructure for Data Centric Analysis and Modelling

Autoren: Ivan Lupelli, Shaun de Witt, Jonathan Hollocombe, David Muir, Rob Akers
Veröffentlicht in: Fusion Engineering and Design, 2016
Herausgeber: 29th Symposium on Fusion Technology

Robust Query Processing in Co-Processor-accelerated Databasesbust Query Processing in Co-Processor-accelerated Databases.

Autoren: Sebastian Breß, Henning Funke, and Jens Teubner
Veröffentlicht in: SIGMOD, 2015
Herausgeber: SIGMOD

Idle Period Propagation in Message-Passing Applications

Autoren: Peng, I. B., Markidis, S., Laure, E., Kestor, G., & Gioiosa, R
Veröffentlicht in: HPCC 2016, 2016
Herausgeber: HPCC 2016

Extending message passing interface windows to storage

Autoren: Rivas-Gomez, S., Peng, I.B., Markidis, S., Laure, E., Kestor, G., Gioiosa, R. .
Veröffentlicht in: CCGrid2017, 2017
Herausgeber: CCGrid2017

Leistungen

Veröffentlichungen

Diese Seite teilen

Herunterladen