IXPUG2016
Experts from around the world met at Argonne National Laboratory for the IXPUG 2016 Annual US Meeting (IXPUG2016). They shared experiences with Xeon Phi-based systems, and learned how to optimize software for manycore machines.
Meeting Highlights:
- Monday - Tutorials on MPI and OpenMP, Programming with Xeon Phi
- Tuesday - Plenary Talks by HPC centers on Deployment and Early Science, Keynotes by Rick Stevens (Argonne) and Joe Curley (Intel)
- Wednesday - Technical Talks on Applications, Memory Models and Tuning, Tutorials on Optimization
- Thursday - Technical Talks on Tools, Techniques and Visualization, Keynote by Jim Tullos (Intel)
Survey: Meeting attendees, please provide your feedback (update: survey closed post-event)
Agenda:
Monday, September 19, 2016
| Start | End | Title | Author(s) | Presentation | Video |
| 9:00am | 12:00 pm | Advanced API Programming | Pavan Balaji | ||
| 1:00 pm | 3:00 pm | Advanced OpenMP | Lars Koesterke and Kent Milfeld | ||
| 3:00 pm | 3:30 pm | Break | |||
| 3:30 pm | 5:30 pm | Ensuring Program Correctness and Optimal Performance on Intel Phi Using Allinea Forge | Ryan Hulguin and Beau Paisley | ||
| Techniques Tutorial | |||||
| 9:00 am | 12:00 pm | Vectorization Strategies for Intel’s 2nd Generation Intel Xeon® Phi™ Architecture Codenamed Knights Landing | Ashish Jha, Vitali Morozov and Jack Deslippe | ||
| 1:00 pm | 5:00 pm | Top Ten Issues for Effectively Utilizing a Large KNL System | John Levesque | ||
Tuesday, September 20, 2016
| Start | End | Title | Author(s) | Presentation | Video |
| 9:00 am | 9:10 am | Welcome from the IXPUG2016 Chair | David Martin | ||
| 9:10 am | 9:20 am | Welcome from the IXPUG Chair | Richard Gerber | Video | |
| 9:20 am | 10:00 am | Backwards and Forwards: The Road to Many Core Computing | Joe Curley | Video | |
| 10:00 am | 10:30 am | ALCF: Overview and Future Thoughts | Mike Papka | ||
| 11:00 am | 12:15 pm | Major Sites Updates: NERSC, TACC, ALCF |
Richard Gerber, John Cazes and Kalyan Kumaran |
||
| 1:30 pm | 2:30 pm | DOE and NIH Partnerships: Cancer and Brain | Rick Stevens | ||
| 2:30 pm | 3:45 pm | Early Science with KNL at TACC, ALCF and NERSC | Jack Deslippe, Tim Williams and Lars Koesterke | Video | |
| 4:15 pm | 4:45 pm | OpenFOAM Optimization on KNL | Prasad Pawar, Paul Edwards, Nishant Agrawal, Ravi Ojha and Sonia Rani | Video | |
| 4:45 pm | 5:15 pm | Optimizing Particle-In-Cell Codes for Intel Xeon Phi Processors | Mathieu Lobet, Henri Vincenti, Jean-Luc Vay and Jack Deslippe | Video |
Wednesday, September 21, 2016
| Start | End | Title | Author(s) | Presentation | Video |
| 8:45 am | 9:15 am | Optimizations of Bspline-based Orbital Evaluations in Quantum Monte Carlo on Multi/Many-core Shared Memory Processors | Ye Luo, Amrita Mathuriya, Anouar Benali, Luke Shulenburger and Jeongnim Kim | Video | |
| 9:15 am | 9:45 am | Enabling Large-Scale Hybrid Density Functional Theory Based Ab Initio Molecular Dynamics in Condensed-Phase Systems | Robert Distasio, Junteng Jia and Alvaro Vazquez-Mayagoitia | Video | |
| 9:45 am | 10:15 am | Performance Optimization of Quantum Espresso on KNL | Taylor Barnes and Jack Deslippe | Video | |
| 10:45 am | 11:15 am | Cache Blocking using Tiling in a Molecular Dynamics Application | Benny Mathew and Manoj Nambiar | Video | |
| 11:15 am | 11:45 am | Efficient MPI/OpenMP Parallelization of the Hartree-Fock Method on Intel® Xeon Phi Processors | Yuri Alexeev, Michael D'Mello, Vladimir Mironov and Alexander Moskovsky | Video | |
| 11:45 am | 11:55 am | MILC Staggered Conjugate Gradient Performance on Intel KNL | Ruizi Li, Carleton Detar, Douglas Doerfler, Steven Gottlieb, Ashish Jha, Balint Joo, Dhiraj Kalamkar and Doug Toussaint | Video | |
| 11:55 am | 12:05 pm | Optimizing Magnetic Fusion PIC Code XGC1 for Xeon Phi | Tuomas Koskela | ||
| 12:05 pm | 12:15 pm | Scaling the Performance of a FDFD Geophysical-imaging Application to Multi-node KNL Clusters | Tareq Malas, Thorsten Kurth and Jack Deslippe | ||
| 1:30 pm | 2:00 pm | Accurate and Efficient Earthquake Simulations on Intel Xeon Phi | Josh Tobin, Alexander Breuer, Charles Yount, Alexander Heinecke and Yifeng Cui | Video | |
| 2:00 pm | 2:30 pm | Seismic Simulations with Local Time Stepping on Xeon and Xeon Phi Processors | Alexander Breuer, Alexander Heinecke and Yifeng Cui | Video | |
| 2:30 pm | 3:00 pm | Grid: Structured Cartesian Mesh Library for Quantum Chromodynamics | Peter Boyle, Azusa Yamaguchi, Guido Cossu and Antonin Portelli | Video | |
| 3:30 pm | 3:40 pm | Enabling High-performance Simulation of Subsurface Flows and Geochemical Processes with Chombo-Crunch on Intel Xeon Phi Knights Landing | Andrey Ovsyannikov | Video | |
| 3:40 pm | 3:50 pm | Reconstructing Particle Trajectories in High Energy Physics with Xeon and Xeon Phi | David Abdurachmanov, Peter Elmer, Giuseppe Cerati, Slava Krutelyov, Steven Lantz, Matthieu Lefebvre, Kevin McDermott, Daniel Riley, Matevz Tadel, Peter Wittich, Frank Wuerthwein and Avi Yagil | ||
| 3:50 pm | 4:00 pm | HACC on the KNL - Porting, Optimizing, and Early Experiences | Hal Finkel, Adrian Pope and JD Emberson | ||
| 4:00 pm | 4:30 pm | Optimizing MFDn (a Nuclear Physics CI Code) for KNL | Pieter Maris | Video | |
| 4:30 pm | 5:00 pm | Boundary Element Method for Manycore Architectures | Jan Zapletal and Michal Merta | ||
| Memory Models and Methods | |||||
| 8:45 am | 9:15 am | Managing Per Job Reconfiguration of Xeon Phi Knights Landing Clusters | Michael Hebenstreit | ||
| 9:15 am | 9:45 am | MCDRAM on 2nd Generation Intel® Xeon Phi™ Processor | James Tullos and Karthik Raman | ||
| 9:45 am | 10:15 am | Cache Size on KNL at LANL | Phillip Romero | ||
| 10:45 am | 11:15 am | Chroma on Knights Landing Three Ways | Balint Joo, Frank Winter, Thorsten Kurth and Jacques Bloch | ||
| 11:15 am | 11:45 am | Tuning Generated Code for KNL | James Osborn | ||
| 11:45 am | 11:55 am | Asynchronous Memory Migration on the KNL | Swann Perarnau and Kamil Iskra | ||
| 11:55 am | 12:05 pm | Understanding Knight's Landing HIgh Bandwidth Memory Using the STREAM Benchmark | Agrima Bahl and Brian Austin | ||
| 12:05 pm | 12:15 pm | Bringing the Customer to the Tech | Richard Coffey and Jini Ramprakash | ||
| Programming Tutorial | |||||
| 1:30 pm | 5:00 pm | Programming Intel’s 2nd Generation Xeon Phi (Knights Landing) | Carlos Rosales-Fernandez, Kent Milfeld and John Cazes | ||
| Optimization Tutorial | |||||
| 8:45 am | 12:15 pm | Optimize Your Workload for Memory and SIMD Parallelism on x86 Platforms with Intel® Advisor XE | Zakhar Matveev and Michael D'mello | ||
| 1:30 pm | 5:30 pm | Optimizing Codes Using the Roofline Model | Jack Deslippe, Zakhar Matveev, Richard Gerber, Brandon Cook, Tuomas Koskela, Mathieu Lobet and Tareq Males | ||
Thursday, September 22, 2016
| Programming and Visualization Tutorial | |||||
| Start | End | Title | Author(s) | Presentation | Video |
| 9:00 am | 9:30 am | Heterogeneous Programming of Intel Xeon and Intel Xeon Phi with Intel TBB Flow Graph | Evgeny Fiksman, Michael Voss and Sergey Vinogradov | Video | |
| 9:30 am | 10:00 am | LLVM and Clang on the KNL | Hal Finkel | Video | |
| 10:00 am | 10:10 am | MPI + hStreams in NEMO5: Partitioning Xeon Phi | Xinchen Guo, Daniel Lemus, Daniel Mejia, Jim Fonseca, Gerhard Klimeck and Tillmann Kubis | ||
| 10:10 am | 10:20 am | Process and Thread Affinity with MPI/OpenMP on KNL | Helen He | Video | |
| 10:20 am | 10:30 am | Effective OpenMP Implementations on Intel's Knights Landing | Yuliana Zamora | ||
| 11:00 am | 11:30 am | Visualization with OSPRay: Research and Production | Aaron Knoll and Ingo Wald | Video | |
| 11:30 am | 11:45 am | Visualization with vl3. Now on the CPU | Silvio Rizzi, Joseph Insley and Aaron Knoll | Video | |
| 1:00 pm | 2:00 pm | Performance Tuning on KNL with Intel® Parallel Studio XE 2017 Cluster Edition | James Tullos | Video | |
| Working Group Meeting | |||||
| 2:00 pm | 3:00 pm | Joint Working Group Meetings | John Pennycook and WG members | ||
Printable Agenda for PDFs: Monday, Tuesday, Wednesday, Thursday
Slides: IXPUG Presentation Repository
Video Broadcast: Webcast Schedule
Streaming Video Links: IXPUG Tuesday, IXPUG Wednesday, IXPUG Thursday
IXPUG2016 Program Committee:
| David | Martin (chair) | Argonne National Laboratory |
| Richard | Gerber (vice-chair) | Lawrence Berkeley National Laboratory |
| Gilles | Civario | Dell |
| Ryan | Coleman | Sandia National Laboratories |
| Jeanine | Cook | Sandia National Laboratories |
| Douglas | Doerfler (tutorial chair) | Lawrence Berkeley National Laboratory |
| Melyssa | Fratkin | Texas Advanced Computing Center (TACC) |
| Rahul | Hardikar | Indian Institute of Science Education and Research (IISER), Pune |
| Steve | Leak | Lawrence Berkeley National Laboratory |
| Michael | Lysaght | Irish Centre for High End Computing (ICHEC) |
| Kent | Milfeld | Texas Advanced Computing Center (TACC) |
| John | Pennycook | Intel Corporation |
| Abhinav | Sarje | Lawrence Berkeley National Laboratory |
| Lisa | Smith | Intel Corporation |
| Thomas | Steinke | Zuse Institute Berlin |
| Estela | Suarez | Forschungszentrum Juelich |
| Srinath | Vadlamani | ParaTools, Inc. |
| Jerome | Vienne | Texas Advanced Computing Center (TACC) |
| Georg | Zitzlsberger | Intel Corporation |
Important Dates:
| Abstract Submission Deadline | 18 August 2016 |
| Abstracts reviewed by IXPUG Committee | 19 - 26 August 2016 |
| Acceptance Notification | 29 August 2016 |
|
Final Presentations due from Speakers |
16 September 2016 |
| Final Agenda published | 9 September 2016 |
Meeting Location: Argonne National Laboratory, TCS Conference Center, Argonne, IL 60439. Argonne map: http://www.anl.gov/downloads/map-Argonne
Event Details
