PAVE: Performance Analysis and Visualization at Exascale

Publications

Papers

Abhinav Bhatele, Stephanie Brink, and Todd Gamblin. Hatchet: Pruning the Overgrowth in Parallel Profiles. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘19), November 2019. LLNL-CONF-772402.

Samuel A. Pollard, Nikhil Jain, Stephen Herbein, and Abhinav Bhatele. Evaluation of an Interference-free Node Allocation Policy on Fat-tree Clusters. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘19), November 2019. LLNL-CONF-745526.

Staci Smith, Clara Cromey, David K. Lowenthal, Jens Domke, Nikhil Jain, and Abhinav Bhatele. Mitigating Inter-Job Interference Using Adaptive Flow-Aware Routing. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘18), November 2018. LLNL-CONF-745538.

Nikhil Jain, Abhinav Bhatele, Xiang Ni, Todd Gamblin, and Laxmikant V. Kale. Partitioning Low-diameter Networks to Eliminate Inter-job Interference. In Proceedings of the IEEE International Parallel & Distributed Processing Symposium (IPDPS’17), May 2017. LLNL-CONF-706801.

Nikhil Jain, Abhinav Bhatele, Samuel T. White, and Todd Gamblin, Laxmikant V. Kale. Evaluating HPC Networks via Simulation of Parallel Workloads. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘16), November 2016. LLNL-CONF-690662.

Abhinav Bhatele, Nikhil Jain, Yarden Livnat, and Valerio Pascucci, and Peer-Timo Bremer. Analyzing Network Health and Congestion in Dragonfly-based Systems. In Proceedings of the IEEE International Parallel & Distributed Processing Symposium (IPDPS’16), May 2016. LLNL-CONF-678293.

Abhinav Bhatele, Andres R. Titus, Jayaraman J. Thiagarajan, Nikhil Jain, Todd Gamblin, Peer-Timo Bremer, Martin Schulz, and Laxmikant V. Kale. Identifying the Culprits behind Network Congestion. In Proceedings of the IEEE International Parallel & Distributed Processing Symposium (IPDPS’15), May 2015. LLNL-CONF-663150.

Nikhil Jain, Abhinav Bhatele, Xiang Ni, Nicholas J. Wright,  and Laxmikant V. Kale. Maximizing Throughput on a Dragonfly Network. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘14), November 2014. LLNL-CONF-653557.

Alfredo Gimenez, Todd Gamblin, Barry Rountree, Abhinav Bhatele, Ilir Jusufi, Peer-Timo Bremer, and Bernd Hamann. Dissecting On-Node Memory Access Performance: A Semantic Approach. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘14), November 2014. LLNL-CONF-658626.

Katherine E. Isaacs, Peer-Timo Bremer, Ilir Jusufi, Toed Gamblin, and Abhinav Bhatele, Martin Schulz, and Bernd Hamann. Combing the Communication Hairball: Visualizing Parallel Execution Traces using Logical Time. In Proceedings of IEEE Transactions on Visualization and Computer Graphics, November 2014. LLNL-JRNL-657418.

Abhinav Bhatele,  Nikhil Jain, Katherine E. Isaacs, Ronak Buch, Todd Gamblin, Steven H. Langer, and Laxmikant V. Kale. Improving Application Performance via Task Mapping on IBM Blue Gene/Q. In Proceedings of IEEE International Conference on High Performance Computing, December 2014. LLNL-CONF-655465.

Nikhil Jain, Abhinav Bhatele, Michael Robson, Todd Gamblin, and Laxmikant Kale. Predicting Application Performance Using Supervised Learning on Communication Features. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘13), November 2013. LLNL-CONF-635857.

Bhatele, Abhinav and Mohror, Kathryn and Langer, Steven H. and Isaacs, Katherine E. There Goes the Neighborhood: Performance Degradation Due to Nearby Jobs. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘13), November 2013. LLNL-CONF-635776.

Abhinav Bhatele, Todd Gamblin, Katherine E. Isaacs, Brian T. N. Gunney, Martin Schulz, Peer-Timo Bremer, Bernd Hamann, Novel views of performance data to analyze large-scale adaptive applications, In Proceedings of ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘12), November 2012. LLNL-CONF-554552.

Abhinav Bhatele, Todd Gamblin, Steven H. Langer, Peer-Timo Bremer, Erik W. Draeger, Bernd Hamann, Katherine E. Isaacs, Aaditya G. Landge, Joshua A. Levine, Valerio Pascucci, Martin Schulz, Charles H. Still, Mapping applications with collectives over sub-communicators on torus networks, In Proceedings of ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC ‘12), November 2012. LLNL-CONF-556491.

Aaditya G. Landge, Joshua A. Levine, Katherine E. Isaacs, Abhinav Bhatele, Todd Gamblin, Martin Schulz, Steve H. Langer, Peer-Timo Bremer, and Valerio Pascucci. Visualizing network traffic to understand the performance of massively parallel simulations. IEEE Transactions on Visualization and Computer Graphics, 2012. LLNL-CONF-543359.

Steven Langer, Abhinav Bhatele, Todd Gamblin, Bert Still, Denise Hinkel, Mike Kumbera, Bruce Langdon, Ed Williams, Simulating Laser-Plasma Interaction in Experiments at the National Ignition Facility on a Cray XE6, In Cray User Group Meeting (CUG ‘12), Stuttgart, Germany, April 2012. LLNL-PROC-547711.

Martin Schulz, Abhinav Bhatele, Peer-Timo Bremer, Todd Gamblin, Katherine Isaacs, Joshua Levine, and Valerio Pascucci. Creating a Tool Set for Optimizing Topology-aware Node Mappings. In Proceedings of the 5th ZIH Parallel Tools Workshop, Dresden, Germany, September 2011. LLNL-CONF-402937.

Martin Schulz, Joshua A. Levine, Peer-Timo Bremer, Todd Gamblin, and Valerio Pascucci. Interpreting performance data across intuitive domains. In International Conference on Parallel Processing (ICPP’11), Taipei, Taiwan, September 13-16 2011 LLNL-CONF-476091.

Steven H. Langer, Bert Still, Peer-Timo Bremer, Denise Hinkel, Bruce Langdon, Joshua Levine, and Ed Williams. Cielo Full-System Simulations of Multi-Beam Laser-Plasma Interaction in NIF Experiments. In Cray Users’ Group Meeting (CUG ‘11). Fairbanks, Alaska. 2011. LLNL-PROC-482696.

Technical Posters

Abhinav Bhatele, Todd Gamblin, Steven H. Langer, Peer-Timo Bremer, and Martin Schulz. Mapping collectives over sub-communicators on torus networks. In Current Challenges in Computing 2012: Network Science, Napa, CA, August 2012. LLNL-POST-563791.

Abhinav Bhatele, Todd Gamblin, Martin Schulz, Peer-Timo Bremer, Intuitive Visualizations for Analyzing Exascale Workloads, In Exascale Research Conference, Portland, Oregon. April 2012. LLNL-POST-545412.

Aaditya Landge, Joshua A. Levine, Peer-Timo Bremer, Martin Schulz, Todd Gamblin, Abhinav Bhatele, Katherine E. Isaacs, Valerio Pascucci, Interactive Linked Visualizations for Performance Analysis of Heterogeneous Computing Clusters, In GPU Technology Conference, 2012. LLNL-POST-518831.

Abhinav Bhatele, Todd Gamblin, Brian T. Gunney, Martin Schulz, Peer-Timo Bremer and Katherine E. Isaacs. Revealing Performance Artifacts in Parallel Codes Through Multi-Domain Visualizations. In SIAM Conference on Parallel Processing. Savannah, Georgia. February 2012. LLNL-POST-527971.

Presentations

Abhinav Bhatele, Peer-Timo Bremer, Todd Gamblin, Martin Schulz, PAVE: Intuitive visualizations for analyzing exascale workloads, Exascale Research Conference, Portland, OR. April 2012. LLNL-PRES-540811.

Brian T.N. Gunney, Abhinav Bhatele, and Todd Gamblin. Tree-based communication for scalable mesh adaptation in the SAMRAI framework. 2012 SIAM Annual Meeting (AN ‘12). Minneapolis, Minnesota. July 2012. LLNL-PRES-562671.

Martin Schulz, Abhinav Bhatele, Peer-Timo Bremer, Todd Gamblin, Katherine Isaacs, Aaditya Landge, Joshua Levine and Valerio Pascucci. A Case for More Modular and Intuitive Performance Analysis Tools. SIAM Conference on Parallel Processing. Savannah, Georgia. February 2012. LLNL-PRES-530515

Martin Schulz. More Intuitive Performance Analysis. Invited talk, DOE Office of Science. Germantown, MD, September 2011. LLNL-PRES-497594.

Martin Schulz. More Intuitive Performance Analysis. Invited talk, Institute of Computer Science, Foundation for Research and Technology Hellas (FORTH). Heraklion, Greece. September 2011.  LLNL-PRES-501271.

Martin Schulz. Performance and Optimization: A Case for more Modular and Intuitive Tools. Institute for Nuclear Theory Exascale Workshop, Seattle, WA, June 2011. LLNL-PRES-490045.

Martin Schulz. A Case for More Intuitive Performance Analysis. Salishan Conference on High-Speed Computing, Salishan, OR, April 2011. LLNL-PRES-481656.

Technical Reports

Abhinav Bhatele, Peer-Timo Bremer, Todd Gamblin, Martin Schulz, Intuitive visualizations through multi-domain projections for performance analysis at scale, Exascale Research Conference, Portland, OR. April 2012.
LLNL-TR-537251.

Abhinav Bhatele, Peer-Timo Bremer, Todd Gamblin, Martin Schulz, PAVE: Intuitive visualizations for analyzing exascale workloads, Exascale Research Conference, Portland, OR. April 2012. LLNL-MI-535518 (position paper)