Performance and Energy Usage of Workloads on KNL and Haswell Architectures. In: High Performance Computing Systems. Performance Modeling, Benchmarking, and Simulation

Allen, Tyler; Daley, Christopher S.; Doerfler, Douglas; Austin, Brian; Wright, Nicholas J.

doi:10.1007/978-3-319-72971-8_12

Performance and Energy Usage of Workloads on KNL and Haswell Architectures. In: High Performance Computing Systems. Performance Modeling, Benchmarking, and Simulation

Conference · Mon Jan 01 04:00:00 EST 2018 · Lecture Notes in Computer Science

DOI:https://doi.org/10.1007/978-3-319-72971-8_12· OSTI ID:1546612

Allen, Tyler ^[1]; Daley, Christopher S. ^[2]; Doerfler, Douglas ^[2]; Austin, Brian ^[2]; Wright, Nicholas J. ^[2]

Clemson Univ., SC (United States)
Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)

Manycore architectures are an energy-efficient step towards exascale computing within a constrained power budget. The Intel Knights Landing (KNL) manycore chip is a specific example of this and has seen early adoption by a number of HPC facilities. It is therefore important to understand the performance and energy usage characteristics of KNL. In this paper, we evaluate the performance and energy efficiency of KNL in contrast to the Xeon (Haswell) architecture for applications representative of the workload of users at NERSC. We consider the optimal MPI/OpenMP configuration of each application and use the results to characterize KNL in contrast to Haswell. As well as traditional DDR memory, KNL contains MCDRAM and we also evaluate its efficacy. Our results show that, averaged over our benchmarks, KNL is 1.84 × more energy efficient than Haswell and has 1.27 × greater performance.

Research Organization:: Lawrence Berkeley National Laboratory-National Energy Research Scientific Computing Center (NERSC)

Sponsoring Organization:: USDOE Office of Science (SC)

DOE Contract Number:: AC02-05CH11231

OSTI ID:: 1546612

Conference Information:: Journal Name: Lecture Notes in Computer Science Journal Volume: 10724

Country of Publication:: United States

Language:: English

References (13)

Castro: a new Compressible Astrophysical Solver. i. Hydrodynamics and Self-Gravity Almgren, A. S.; Beckner, V. E.; Bell, J. B. The Astrophysical Journal, Vol. 715, Issue 2 https://doi.org/10.1088/0004-637X/715/2/1221	journal	May 2010
Argonne applications for the IBM Blue Gene/Q, Mira Coghlan, S.; Kumaran, K.; Loy, R. M. IBM Journal of Research and Development, Vol. 57, Issue 1/2 https://doi.org/10.1147/JRD.2013.2238371	journal	January 2013
Effective Holistic Performance Measurement at Petascale Using IPM Fürlinger, Karl; Wright, Nicholas J.; Skinner, David Competence in High Performance Computing 2010 https://doi.org/10.1007/978-3-642-24025-6_2	book	January 2011
Determination of performance characteristics of scientific applications on IBM Blue Gene/Q Evangelinos, C.; Walkup, R. E.; Sachdeva, V. IBM Journal of Research and Development, Vol. 57, Issue 1/2 https://doi.org/10.1147/JRD.2012.2229901	journal	January 2013
Runtime Power Limiting of Parallel Applications on Intel Xeon Phi Processors Lawson, Gary; Sundriyal, Vaibhav; Sosonkina, Masha 2016 4th International Workshop on Energy Efficient Supercomputing (E2SC) https://doi.org/10.1109/E2SC.2016.011	conference	November 2016
The impact of hyper-threading on processor resource utilization in production applications Saini, Subhash; Jin, Haoqiang; Hood, Robert 2011 18th International Conference on High Performance Computing (HiPC) https://doi.org/10.1109/HiPC.2011.6152743	conference	December 2011
Evaluating and Optimizing the NERSC Workload on Knights Landing Barnes, Taylor; Cook, Brandon; Deslippe, Jack 2016 7th International Workshop on Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems (PMBS) https://doi.org/10.1109/PMBS.2016.010	conference	November 2016
Nyx: A MASSIVELY PARALLEL AMR CODE FOR COMPUTATIONAL COSMOLOGY Almgren, Ann S.; Bell, John B.; Lijewski, Mike J. The Astrophysical Journal, Vol. 765, Issue 1 https://doi.org/10.1088/0004-637X/765/1/39	journal	February 2013
Metrics for Energy-Aware Software Optimisation Roberts, Stephen I.; Wright, Steven A.; Fahmy, Suhaib A. Lecture Notes in Computer Science https://doi.org/10.1007/978-3-319-58667-0_22	book	January 2017
Introducing FIRESTARTER: A processor stress test utility Hackenberg, Daniel; Oldenburg, Roland; Molka, Daniel 2013 International Green Computing Conference (IGCC), 2013 International Green Computing Conference Proceedings https://doi.org/10.1109/IGCC.2013.6604507	conference	June 2013
Capability Models for Manycore Memory Systems: A Case-Study with Xeon Phi KNL Ramos, Sabela; Hoefler, Torsten 2017 IEEE International Parallel and Distributed Processing Symposium (IPDPS) https://doi.org/10.1109/IPDPS.2017.30	conference	May 2017
Performance Analysis and Workload Characterization with IPM Fürlinger, Karl; Wright, Nicholas J.; Skinner, David Tools for High Performance Computing 2009 https://doi.org/10.1007/978-3-642-11261-4_3	book	January 2010
Effective Performance Measurement at Petascale Using IPM Fuerlinger, Karl; Wright, Nicholas J.; Skinner, David 2010 IEEE 16th International Conference on Parallel and Distributed Systems (ICPADS) https://doi.org/10.1109/ICPADS.2010.16	conference	December 2010

Similar Records

Electronic Structure Theory Calculations Using Modern Architectures: KNL vs Haswell

Journal Article · Tue Oct 26 00:00:00 EDT 2021 · Journal of Chemical Theory and Computation · OSTI ID:1873417

A Locality-Based Threading Algorithm for the Configuration-Interaction Method

Journal Article · Mon Jul 03 00:00:00 EDT 2017 · IEEE International Symposium on Parallel and Distributed Processing, Workshops and Phd Forum · OSTI ID:1393243

MILC staggered conjugate gradient performance on Intel KNL

Conference · Thu Nov 03 00:00:00 EDT 2016 · Proceedings of Science (POS) · OSTI ID:1398438

Related Subjects

Benchmarking
Energy
Haswell
Hyperthreads
Intel Knights Landing
Manycore architecture
Power consumption

Performance and Energy Usage of Workloads on KNL and Haswell Architectures. In: High Performance Computing Systems. Performance Modeling, Benchmarking, and Simulation

Citation Formats

References (13)

Similar Records

Related Subjects