News and Announcements

Reaching No. 1 at the Summit

Image courtesy of the US Department of Energy’s Oak Ridge National Laboratory.

On Friday, June 8th, the US Department of Energy’s Oak Ridge National Laboratory (ORNL) announced that Summit is now the fastest supercomputer in the world, besting China’s Sunway TaihuLight machine and ending a four-year reign of Chinese systems at No. 1. The TOP500 results will be released later this month at ISC 2018, but all indications are that Summit will officially take the No. 1 position in the June 2018 list at ~200 petaFLOP/s.

For some perspective, Summit is roughly eight times faster than its predecessor, Titan, which achieved 27 petaFLOP/s in 2012. Summit is an IBM AC922 heterogeneous computing system consisting of 4,608 compute nodes, with each node containing two 22-core IBM Power9 CPUs and six NVIDIA Tesla V100 GPU accelerators. These components communicate over a dual-rail Mellanox EDR 100 Gbit/s InfiniBand interconnect. More than 10 petabytes of available memory are paired with fast, high-bandwidth pathways for efficient data movement.

US Secretary of Energy, Rick Perry, had this to say about Summit: “Today’s launch of the Summit supercomputer demonstrates the strength of American leadership in scientific innovation and technology development. It’s going to have a profound impact in energy research, scientific discovery, economic competitiveness and national security.” Secretary Perry added, “I am truly excited by the potential of Summit, as it moves the nation one step closer to the goal of delivering an exascale supercomputing system by 2021. Summit will empower scientists to address a wide range of new challenges, accelerate discovery, spur innovation and above all, benefit the American people.”

To read more about Summit, check out this article from the New York Times.

The editor would like to thank Morgan McCorkle (ORNL) for her contributions to this article.

SIGHPC in East TN

The Association for Computing Machinery’s (ACM’s) special interest group in HPC (SIGHPC) has several virtual chapters that focus on different topics (e.g., education, big data). However, given the concentration of people in the East Tennessee region interested in HPC, a small group of ACM members from ORNL and UTK is—with the encouragement of the ACM—forming the first geographically localized chapter of SIGHPC.

The initial stakeholder meeting was held at Calhoun’s on Wednesday, May 23rd, where the Appalachian HPC Organizing Committee was formed as a small ad-hoc group to push forward with the charter details. Over 30 very enthusiastic participants attended the meeting and ironed out three “tent pole” ideas for the regional chapter:

1. What’s in the air?

  • Sharing knowledge and technical context among members.
  • Activities: Flash talks, regional conference, vendor presentations, topical debates (“What happens to HPC in the age of IoT?”), unconference, liaisons to the virtual chapters of SIGHPC (e.g., HPC Education, Big Data and HPC), pre- or post-conference talks.

2. Workforce development.

  • Helping share the excitement of HPC with students in the pipeline as well as professionals in related spheres.
  • Activities: Hackathons, sponsorship of student cluster challenge teams for SC, training sessions, managing speaker list to go to colleges, mentoring.

3. Community building.

  • Working together to build professional and personal rapport.
  • Activities: Awards committee, social meet-up-style networking, shared web presence, recognition.

If you have any interest in joining this regional chapter or would like to volunteer your time to help push the chapter forward, please contact Matthew Wolf.

The editor would also like to remind readers that SIGHPC is enthusiastically endorsed by Jack Dongarra (ca. 2012).

Employment Opportunities at ICL

Research Scientist (with MS or PhD) or Postdoctoral Researcher

ICL is seeking full-time scientists (MS or PhD) or postdoctoral researchers to participate in the design, development, and maintenance of numerical software libraries for solving linear algebra problems on large, distributed-memory machines with multi-core processors, hardware accelerators, and performance monitoring capabilities for new and advanced hardware and software technologies. The prospective researcher will coauthor papers to document research findings, present the team’s work at conferences and workshops, and help lead students and other team members in their research endeavors in ongoing and future projects. Given the nature of the work, there will be opportunities for publication, travel, and high-profile professional networking and collaboration across academia, labs, and industry.

An MS or PhD in computer science, computational sciences, or math is preferred.  Background in at least one of the following areas is also preferred: numerical linear algebra, HPC, performance monitoring, machine learning, or data analytics. Full-time employment for up to 4 years with the possibility of further extensions based on funding availability and performance.

Joining this team will offer qualified candidates exciting career opportunities as they participate in the US Department of Energy’s Exascale Computing Project (ECP). ICL is involved in several ECP projects, including SLATE (http://icl.utk.edu/slate/), PEEKS (http://icl.utk.edu/peeks/), xSDK (http://www.icl.utk.edu/research/xsdk4ecp), Exa-PAPI (http://icl.utk.edu/exa-papi/), CEED (https://ceed.exascaleproject.org/), Distributed Tasking for Exascale (PaRSEC) (http://icl.utk.edu/dte/), MAGMA (http://icl.cs.utk.edu/magma/), FFT-ECP, and others.

Starting date is July 1, 2018 or later. All qualified candidates, be it fresh (MS or PhD) graduates or seasoned HPC veterans, are encouraged to apply.

For more information, contact Jack Dongarra (dongarra@icl.utk.edu) or check out ICL’s jobs page: http://www.icl.utk.edu/jobs.

Conference Reports

IPDPS 2018

On May 21–25, ICL’s Ichitaro Yamazaki and Jakub Kurzak—along with hundreds of other attendees—made their way to Vancouver, British Columbia for the IEEE International Parallel & Distributed Processing Symposium (IPDPS). IPDPS is an international forum that enables engineers and scientists from around the world to present their latest research findings in all aspects of parallel computing. In addition to technical sessions and paper presentations, the meeting offers workshops, tutorials, and commercial presentations and exhibits.

Ichi presented a paper that he coauthored with Ahmad, Stan, and Jack—and our collaborators from the Joint Usage/Research Center for Interdisciplinary Large-scale Information Infrastructures (JHPCN), Ida, Ohshima, and Yokota—about using MAGMA’s variable-size batched kernel to accelerate hierarchical matrix computation on GPUs.

Jakub served as Program Chair for the international Workshop on Automatic Performance Tuning (iWAPT). The goal of iWAPT is to bring together researchers who are investigating automated techniques for constructing and/or adapting algorithms and software for high-performance on modern complex machine architectures.

The editor would like to thank Ichitaro Yamazaki and Jakub Kurzak for their contributions to this article.

Interview

John Batson Then

John Batson

Where are you from, originally?
I’m from Knoxville, Tennessee, and I’ve been here my whole life.

Can you summarize your educational background?
My BA is in technical communication from UTK. More by chance than by design, most of my coursework in this discipline was “experiential learning,” which involved working with non-native users of English on their scientific and other technical projects. I also took quite a few literature courses, where my professors urged me to go to grad school. Starting in August, I’ll pursue my MA in the literature, criticism, and textual studies program at UTK.

How did you first hear about ICL, and what made you want to work here?
I heard about ICL from Dr. Russel Hirst, my undergraduate advisor. I was aware of several summer internship opportunities, but none of them really interested me. When ICL’s Sam Crawford told Dr. Hirst about an internship opportunity here, however, I applied almost immediately. I was amazed by the work that goes on here—and amazed that I hadn’t been aware of it. I wanted to contribute.

What is your focus here? What are you working on?
I’m primarily helping Sam edit research papers before submission, as revisions are being made, and so on. I’ve also helped polish the language on a couple of posters and sponsor reports. My biggest project so far (also my first) was one of the SLATE Working Notes. Sometimes I discuss semantics with Anthony.

What would you consider your most valuable “lesson” you have learned so far at ICL?
I feel I’ve learned a number of valuable lessons here. Principal among them is that if you don’t understand something, you can just ask someone who does. That may seem obvious, platitudinous even, but I think a lot people my age fail to speak up for fear of embarrassment or looking stupid at their first job post-college. I’ve heard horror stories of people sitting frozen in their offices for hours because they’re afraid to ask for help. You can just ask.

Along the same lines, I’ve learned the value of having a good mentor. I don’t mean to seem like a flatterer, but if Sam weren’t receptive to emails and knocks on his door, as a lifelong Windows user I’d probably still be trying to figure out how to use my Mac. On a more technical level, I’m learning to use LaTeX. Also, I’ve learned to close my door when Anthony wants to discuss semantics.

What are your interests/hobbies outside of work?
I’m big into music—all types of music. My friend has what is essentially a mini music studio in his basement, and because nobody else wanted to do it, I started taking lessons from a vocal coach a few years ago. My friends and I formed a “band,” which I put in quotes because we don’t play shows, and we’re terrible. It’s fun, though.

Tell us something about yourself that might surprise people.
I won a national championship at a tennis tournament when I was 17.

If you weren’t working at ICL, where would you like to be working and why?
I’d probably be trying to find a job as an editor for an academic publication. I work a few hours a week doing just that for the International Journal of Nuclear Security, a publication of the Institute for Nuclear Security here at UTK. I like engaging with important work, so whether I’m doing that while helping researchers at a lab like ICL or with a publication, I’ll be content.

Recent Papers

  1. Haidar, A., A. Abdelfattah, M. Zounon, S. Tomov, and J. Dongarra, A Guide for Achieving High Performance with Very Small Matrices on GPUs: A Case Study of Batched LU and Cholesky Factorizations,” IEEE Transactions on Parallel and Distributed Systems, vol. 29, issue 5, pp. 973–984, May 2018.  (832.92 KB)
  2. Dong, T., A. Haidar, S. Tomov, and J. Dongarra, Accelerating the SVD Bi-Diagonalization of a Batch of Small Matrices using GPUs,” Journal of Computational Science, vol. 26, pp. 237–245, May 2018.  (2.18 MB)
  3. Gates, M., S. Tomov, and J. Dongarra, Accelerating the SVD Two Stage Bidiagonal Reduction and Divide and Conquer Using GPUs,” Parallel Computing, vol. 74, pp. 3–18, May 2018.  (1.34 MB)
  4. Yamazaki, I., A. Abdelfattah, A. Ida, S. Ohshima, S. Tomov, R. Yokota, and J. Dongarra, Analyzing Performance of BiCGStab with Hierarchical Matrix on GPU Clusters,” IEEE International Parallel and Distributed Processing Symposium (IPDPS), Vancouver, BC, Canada, IEEE, May 2018.  (1.37 MB)
  5. Abdelfattah, A., A. Haidar, S. Tomov, and J. Dongarra, Batched One-Sided Factorizations of Tiny Matrices Using GPUs: Challenges and Countermeasures,” Journal of Computational Science, vol. 26, pp. 226–236, May 2018.  (3.73 MB)
  6. Caniou, Y., E. Caron, A K W. Chang, and Y. Robert, Budget-Aware Scheduling Algorithms for Scientific Workflows with Stochastic Task Weights on Heterogeneous IaaS Cloud Platforms,” 2018 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), Vancouver, BC, Canada, IEEE, May 2018.  (1.31 MB)
  7. Bouteiller, A., G. Bosilca, T. Herault, and J. Dongarra, Data Movement Interfaces to Support Dataflow Runtimes,” Innovative Computing Laboratory Technical Report, no. ICL-UT-18-03: University of Tennessee, May 2018.  (210.94 KB)
  8. Jagode, H., A. Danalis, R. Hoque, M. Faverge, and J. Dongarra, Evaluation of Dataflow Programming Models for Electronic Structure Theory,” Concurrency and Computation: Practice and Experience: Special Issue on Parallel and Distributed Algorithms, vol. 2018, issue e4490, pp. 1–20, May 2018.  (1.69 MB)
  9. Herault, T., Y. Robert, A. Bouteiller, D. Arnold, K. Ferreira, G. Bosilca, and J. Dongarra, Optimal Cooperative Checkpointing for Shared High-Performance Computing Platforms,” 2018 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), Best Paper Award, Vancouver, BC, Canada, IEEE, May 2018.  (899.3 KB)
  10. Luo, X., W. Wu, G. Bosilca, T. Patinyasakdikul, L. Wang, and J. Dongarra, ADAPT: An Event-Based Adaptive Collective Communication Framework,” The 27th International Symposium on High-Performance Parallel and Distributed Computing (HPDC '18), Tempe, Arizona, ACM Press, June 2018.  (493.65 KB)
  11. Bosilca, G., A. Bouteiller, T. Herault, V. Le Fèvre, Y. Robert, and J. Dongarra, Distributed Termination Detection for HPC Task-Based Environments,” Innovative Computing Laboratory Technical Report, no. ICL-UT-18-14: University of Tennessee, June 2018.
  12. Abdelfattah, A., M. Gates, J. Kurzak, P. Luszczek, and J. Dongarra, Implementation of the C++ API for Batch BLAS,” SLATE Working Notes, no. 07, ICL-UT-18-04: Innovative Computing Laboratory, University of Tennessee, June 2018.  (1.07 MB)
  13. YarKhan, A., G. Ragghianti, J. Dongarra, M. Cawkwell, D. Perez, and A. Voter, Initial Integration and Evaluation of SLATE Parallel BLAS in LATTE,” Innovative Computing Laboratory Technical Report, no. ICL-UT-18-07: Innovative Computing Laboratory, University of Tennessee, June 2018.  (366.6 KB)
  14. Kurzak, J., M. Gates, A. YarKhan, I. Yamazaki, P. Luszczek, J. Finney, and J. Dongarra, Parallel Norms Performance Report,” SLATE Working Notes, no. 06, ICL-UT-18-06: Innovative Computing Laboratory, University of Tennessee, June 2018.  (1.13 MB)
  15. Anzt, H., I. Yamazaki, M. Hoemmen, E. Boman, and J. Dongarra, Solver Interface & Performance on Cori,” Innovative Computing Laboratory Technical Report, no. ICL-UT-18-05: University of Tennessee, June 2018.  (188.05 KB)
  16. Haidar, A., A. Abdelfattah, M. Zounon, P. Wu, S. Pranesh, S. Tomov, and J. Dongarra, The Design of Fast and Energy-Efficient Linear Solvers: On the Potential of Half-Precision Arithmetic and Iterative Refinement Techniques,” International Conference on Computational Science (ICCS 2018), vol. 10860, Wuxi, China, Springer, pp. 586–600, June 2018.  (487.88 KB)
  17. Haidar, A., S. Tomov, A. Abdelfattah, M. Zounon, and J. Dongarra, Using GPU FP16 Tensor Cores Arithmetic to Accelerate Mixed-Precision Iterative Refinement Solvers and Reduce Energy Consumption,” ISC High Performance (ISC'18), Best Poster, Frankfurt, Germany, June 2018.  (3.01 MB)
  18. Haidar, A., S. Tomov, A. Abdelfattah, M. Zounon, and J. Dongarra, Using GPU FP16 Tensor Cores Arithmetic to Accelerate Mixed-Precision Iterative Refinement Solvers and Reduce Energy Consumption , Frankfurt, Germany, ISC High Performance (ISC18), Best Poster Award, June 2018.  (3.01 MB)

Recent Conferences

  1. MAY
    -
    IPDPS 2018 Vancouver
    Ichitaro Yamazaki
    Ichitaro
    Jakub Kurzak
    Jakub
    Ichitaro Yamazaki, Jakub Kurzak
  2. MAY
    Stanimire Tomov
    Stan
    Stanimire Tomov

Upcoming Conferences

  1. JUN
    -
    GPU Hackathon Boulder
    Piotr Luszczek
    Piotr
    Piotr Luszczek
  2. JUN
    -
    Heike Jagode
    Heike
    Heike Jagode
  3. JUN
    -
    Gerald Ragghianti
    Gerald
    Jack Dongarra
    Jack
    Gerald Ragghianti, Jack Dongarra
  4. JUN
    -
    HPDC Tempe
    Xi Luo
    Xi
    Xi Luo
  5. JUN
    -
    Thomas Herault
    Thomas
    Thomas Herault
  6. JUN
    -
    Piotr Luszczek
    Piotr
    Piotr Luszczek
  7. JUN
    -
    George Bosilca
    George
    George Bosilca

Recent Lunch Talks

  1. MAY
    4
    Ahmad Abdelfattah
    Ahmad Abdelfattah
    MAGMA Update: High Performance and Energy Efficient LU Factorization PDF
  2. MAY
    11
    Ana Gainaru
    Ana Gainaru from Vanderbilt University
    Scheduling Solutions for Data-Driven Large-Scale Applications PDF
  3. MAY
    18
    Pratik Nayak
    Pratik Nayak from Karlsruhe Institute of Technology
    Using Iterative Methods for Local Solves in Asynchronous Schwarz Methods
  4. MAY
    25
    Yaohung Tsai
    Yaohung Tsai
    Pseudo-Assembly Programming on NVIDIA GPU PDF

Upcoming Lunch Talks

  1. JUN
    1
    Anne Benoit
    Anne Benoit from Georgia Tech
    Combining Checkpointing and Replication for Reliable Execution of Linear Workflows
  2. JUN
    8
    Lou Gross
    Lou Gross from NIMBioS
    A Rational Basis for Hope: Human Behavior Modeling and Climate Change

People

  1. Daniel Barry
    Daniel Barry joined ICL this summer as a Graduate Research Assistant. Also a student at the Bredesen Center, Daniel will be working with the PICL team. Welcome aboard, Daniel.
  2. Maksims Abalenkovs
    Maksims Abalenkovs from the University of Manchester is visiting ICL from May 28th to June 8th and will be working with Jakub and the LA group.

Visitors

  1. Artur Safin
    Artur Safin from University of Texas at Dallas will be visiting from June 3 through June 4. To interview for position
  2. Hejer Shaiek
    Hejer Shaiek from ENSEEIHT will be visiting from June 5 through August 31. Hejer is working with Stan on the MAGMA project.
  3. Yuechao Lu
    Yuechao Lu from Osaka University will be visiting from June 27 through December 18. Intern; working with the MAGMA team

congratulations

Jackson David Bachor

The Bachor and Dongarra families recently welcomed a new baby boy into the fold. Making a surprise arrival a few weeks early, but stout and healthy nevertheless, Jackson David Bachor was born on May 29, 2018 to proud parents Katie and Ryan Bachor and proud grandparents Sue and Jack Dongarra. Welcome, Jackson!

Dates to Remember

ICL Retreat

The 2018 ICL retreat has been set for August 20–21 at the RT Lodge. Mark your calendars!