ICL Hosts Sparse BLAS Workshop

Sparse BLAS Workshop Group Photo

ICL hosted a Sparse BLAS Workshop November 7-9, 2023. The workshop brought together scientists from major research institutions and industry partners, including AMD, ARM, Intel, MathWorks, and NVIDIA, and focused on developing a common understanding and interface for basic operations on sparse linear algebra objects. For three days, the attendees listened to pitch talks, discussed in breakout sessions, and engaged in hackathons focused on implementing a reference for sparse BLAS operations. The dry workshop sessions were complemented by a social program allowing the participants to explore Knoxville’s downtown cuisine and enjoy a college party. The positive feedback at Supercomputing encourages the participants to continue the effort remotely and plan for the next workshop edition. A special thanks to ICL’s Operations team that ensured both the smooth workshop execution and that nobody went home hungry.

Treece Burgess Provides Jupyter Notebook Tutorial

ICL’s Treece Burgess recently presented a comprehensive tutorial on the use of Jupyter Notebook. The tutorial is available to view online.

Congratulations

ICL’s Daniel Mishler Awarded with Best Student Poster at IEEE Cluster 2023

IEEE Cluster 2023 was held in Santa Fe, NM October 31st through November 3rd. ICL’s Daniel Mishler brought home the award for Best Student Poster for his research poster Performance Insights into Device-initiated RMA using Kokkos Remote Spaces.

Conference Reports

SC23

This slideshow requires JavaScript.

The International Conference for High Performance Computing, Networking, Storage, and Analysis returned to Denver in 2023, November 12th – 17th. This year’s conference General Chair was ICL alumnus Dorian Arnold.

ICL was once again well represented on the conference program. The ICL@SC23 web page details ICL’s attendees and a full schedule of ICL participation in the workshops, tutorials, and BoFs. Updated project handouts were on hand in the University of Tennessee’s booth space, which also offered conference attendees a place to take a break from collecting swag and enjoy a friendly game of spikeball.

SC23 BBLAS SC23 DPLASMA SC23 Ginkgo SC23 heFFTe
SC23 HPCG SC23 HPL SC23 HPL-MxP SC23 MAGMA
SC23 PAPI SC23 PaRSEC SC23 SLATE SC23 TTG
SC23 ULFM

ICL Alumni Dinner

This slideshow requires JavaScript.

The annual ICL Alumni Dinner was held at Henry’s Tavern in Denver on November 15th, and gathered around 40 past and present ICL members. It was a vibrant evening fostering connections and re-connections among attendees, embodying the collaborative spirit fundamental to ICL’s legacy.

Recent Releases

Ginkgo 1.7.0

The Ginkgo team is proud to announce the new Ginkgo minor release 1.7.0 for its sparse linear algebra package: https://github.com/ginkgo-project/ginkgo/releases/tag/v1.7.0.

This release brings new features such as:

  • Complete GPU-resident sparse direct solvers feature set and interfaces,
  • Improved Cholesky factorization performance,
  • A new MC64 reordering,
  • Batched iterative solver support with the BiCGSTAB solver with batched Dense and ELL matrix types,
  • MPI support for the SYCL backend,
  • Improved ParILU(T)/ParIC(T) preconditioner convergence
  • and more!

LAPACK 3.12.0

The LAPACK team announced version 3.12.0 of the venerable software library. New features include:

  • Dynamic Mode Decomposition (DMD)
  • Truncated QR with Column Pivoting
  • Reciprocal Scaling of complex vectors
  • Improvements in the 64-bit API
  • as well as bug fixes and other changes as detailed at: https://www.netlib.org/lapack/lapack-3.12.0.html

SLATE 2023.11.05

SLATE version 2023.11.05 adds:

  • Updated examples for revised Users and Developers Guides
  • Improved performance using 2D GPU grids and GPU-to-GPU copies
  • Improved error handling for singular matrices
  • Fixed numerous bugs and refactored for better sustainability
  • And more! See: https://github.com/icl-utk-edu/slate/releases/tag/v2023.11.05 for further details.

Interview

You Wu

Undergraduate Research Assistant
ICL Alumnus
You Wu Then

Where are you from originally?

I was born and raised in Xiamen, a subtropical island located on the southeast coast of China, overlooking the Taiwan Strait. Some of my best memories are just hanging out on the beach, walking by the sea with my mom, and playing fetch with my dog Nico.

Can you summarize your educational and professional background?

I completed my high school in Wiesbaden, Germany in 2020. Later that year, I moved to Switzerland to begin my undergrad studies in Computational Science and Engineering (CSE) at ETH Zurich. It was during this time that I discovered an interest in high-performance computing. Joining the student cluster competition team ‘RACKlette’ was a significant step, and it took me all the way to competing at the student cluster competition in Dallas during SC22.

How did you first hear about ICL, and what made you want to work here?

I first heard about ICL when I met Phuong at SC22. She also participated in SCC and she was so passionate about her work that I thought it must be a good place where I could learn a lot!

What are your main research interests and what do you work on at ICL?

I am interested in physically based simulations and particularly in the mechanics of porous media. I also have grown a keen interest in sparse linear algebra during my stay at ICL.

At ICL, I am currently working on improving the language interoperability of the Ginkgo library by creating C bindings and using them for the first prototype of the Julia package Ginkgo.jl. This can help in enhancing Ginkgo’s accessibility for domain scientists.

What would you like to do next in your career?

I would like to further contribute to Ginkgo while learning more about heterogeneous computing. Afterwards, I wish to continue pursuing a Ph.D. after I finish my studies at ETH Zurich.

What are your interests/hobbies outside of work?

I don’t have specific hobbies that I engage in outside of work regularly and I’m often quite content with just sitting idle. But every now and then, I might also feel a pulse to create something different, like carving a seal or doing some traditional Chinese ink-wash painting. Being alone in nature is super relaxing for me too, I enjoy soaking in the peace a lot.

Tell us something about yourself that might surprise people.

My full name Wu You (吴莜) including my surname in its original order means ‘carefree’ in Chinese.

Recent Papers

  1. Valeev, E. F., R. J. Harrison, A. A. Holmes, C. C. Peterson, and D. A. Penchoff, Direct Determination of Optimal Real-Space Orbitals for Correlated Electronic Structure of Molecules,” Journal of Chemical Theory and Computation, vol. 19, issue 20, pp. 7230 - 7241, October 2023.
  2. Li, J., G. Bosilca, A. Bouteiller, and B. Nicolae, Elastic deep learning through resilient collective operations,” SC-W 2023: Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis, Denver, CO, ACM, November 2023.
  3. Abdelfattah, A., S. Tomov, P. Luszczek, H. Anzt, and J. Dongarra, GPU-based LU Factorization and Solve on Batches of Matrices with Band Structure,” SC-W 2023: Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis, Denver, CO, ACM, November 2023.
  4. Ribizel, T., and H. Anzt, Parallel Symbolic Cholesky Factorization,” SC-W 2023: Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis, Denver, CO, ACM, November 2023.
  5. Mishler, D., J. Ciesko, S. Olivier, and G. Bosilca, Performance Insights into Device-initiated RMA Using Kokkos Remote Spaces,” 2023 IEEE International Conference on Cluster Computing Workshops (CLUSTER Workshops), Santa Fe, NM, USA, IEEE, November 2023.
  6. Cao, Q., S. Abdulah, H. Ltaief, M. G. Genton, D. Keyes, and G. Bosilca, Reducing Data Motion and Energy Consumption of Geospatial Modeling Applications Using Automated Precision Conversion,” 2023 IEEE International Conference on Cluster Computing (CLUSTER), Santa Fe, NM, USA, IEEE, November 2023.
  7. Sukkari, D., M. Gates, M. Al Farhan, H. Anzt, and J. Dongarra, Task-Based Polar Decomposition Using SLATE on Massively Parallel Systems with Hardware Accelerators,” SC-W '23: Proceedings of the SC '23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis, Denver, CO, ACM, November 2023.
  8. Slattery, S. A., K. A. Surjuse, C. Peterson, D. A. Penchoff, and E. Valeev, Economical Quasi-Newton Unitary Optimization of Electronic Orbitals,” Physical Chemistry Chemical Physics, December 2023, 2024.
  9. Lindquist, N., P. Luszczek, and J. Dongarra, Generalizing Random Butterfly Transforms to Arbitrary Matrix Sizes : arXiv, December 2023.

Recent Conferences

  1. OCT
    -
    Ahmad Abdelfattah
    Ahmad
    Ahmad Abdelfattah
  2. OCT
    -
    Piotr Luszczek
    Piotr
    Piotr Luszczek
  3. OCT
    -
    EPEXA Project Meeting New York City
    George Bosilca
    George
    Joseph Schuchart
    Joseph
    Thomas Herault
    Thomas
    George Bosilca, Joseph Schuchart, Thomas Herault
  4. OCT
    -
    Invited speaker at the University of Central Florida Orlando
    Deborah Penchoff
    Deborah
    Deborah Penchoff
  5. OCT
    -
    Deborah Penchoff
    Deborah
    Deborah Penchoff
  6. OCT
    -
    ORNLAIWS Atlanta
    Hartwig Anzt
    Hartwig
    Hartwig Anzt
  7. OCT
    -
    Daniel Mishler
    Daniel
    George Bosilca
    George
    Daniel Mishler, George Bosilca
  8. OCT
    -
    Piotr Luszczek
    Piotr
    Piotr Luszczek
  9. NOV
    -
    SC23 Denver
    Ahmad Abdelfattah
    Ahmad
    Anthony Danalis
    Anthony
    Aurelien Bouteiller
    Aurelien
    Daniel Barry
    Daniel
    Deborah Penchoff
    Deborah
    George Bosilca
    George
    Jack Dongarra
    Jack
    Joseph Schuchart
    Joseph
    Mark Gates
    Mark
    Piotr Luszczek
    Piotr
    Thomas Herault
    Thomas
    Thomas Grützmacher
    Thomas
    You Wu
    You
    Ahmad Abdelfattah, Anthony Danalis, Aurelien Bouteiller, Daniel Barry, Deborah Penchoff, George Bosilca, Jack Dongarra, Joseph Schuchart, Mark Gates, Piotr Luszczek, Thomas Herault, Thomas Grützmacher, You Wu

Upcoming Conferences

  1. DEC
    -
    Aurelien Bouteiller
    Aurelien
    Aurelien Bouteiller
  2. DEC
    -
    Joseph Schuchart
    Joseph
    Joseph Schuchart

Recent Lunch Talks

  1. OCT
    6
    Fabrizio Bisetti
    Fabrizio Bisetti from University of Texas at Austin
    Jacobian-free Newton-Krylov method for the simulation of non-thermal plasma discharges with high-order time integration and physics-based preconditioning PDF
  2. OCT
    13
    Thomas Gillis
    Thomas Gillis from ANL
    Enabling PDE simulation at an extreme scale
  3. OCT
    20
    Cory Hauck
    Cory Hauck from ORNL
    Iterative strategies for kinetic models of particle transport PDF
  4. OCT
    27
    Yaohung (Mike) Tsai
    Yaohung (Mike) Tsai from Meta
    MTIA: First Generation Silicon Targeting Meta's Recommendation Systems
  5. NOV
    3
    Anustuv Pal
    Anustuv Pal
    cyPAPI: Enabling hardware performance measurement in Python applications PDF
  6. NOV
    10
    Tobias Ribizel
    Tobias Ribizel from Karlsruhe Institute of Technology
    Building a GPU-resident sparse direct solver from scratch PDF

Upcoming Lunch Talks

  1. DEC
    8
    Yves Robert
    Yves Robert from ENS-Lyon
    Variable-capacity scheduling PDF
  2. DEC
    15
    Mark Gates
    Mark Gates
    Polar Decomposition Using SLATE PDF

Around the Web

@TheRegister

@thoefler

@ScienceMagazine

@mnoukhiya