News and Announcements

25 Years of Innovative Computing

25-Years

As many of you know, the 2014-2015 academic year marks ICL’s 25th year of Innovative Computing. To commemorate this milestone, ICL is hosting a 25th anniversary workshop on April 1 – 2, with a welcome reception on March 31 and a banquet on April 1. The workshop will feature presentations from our current and former group members, and the reception and banquet are open to family and spouses.

Although we do not have an agenda at this time, we have a rough schedule to assist in planning.

Draft schedule
March 31, 2015 Welcome Reception, Downtown Hilton, 6pm
April 1-2 25 years of Innovative Computing, UT Conference Center, Henley Street, 9am-5pm
April 1 Banquet celebrating 25 years, UT’s Neyland Stadium Skybox, 6pm

If you wish to attend, please fill out the registration page no later than February 14, 2015. Keep in mind that ICL will pay for up to 3 nights of lodging from March 31 – April 2, and you also have the option of extending your stay with the Hilton at the state employee rate of $88/night (+ taxes). If you require a letter of invitation, please let Tracy Rafferty know as soon as possible (rafferty@icl.utk.edu).

If you are unable to attend, please go to the RSVP page and update your contact information for our records. Please note: Campus has changed, even in the past year, so we will be providing maps to relevant venues and areas of campus.

More information available at http://icl.utk.edu/25, and we look forward to seeing you all!

Conference Reports

Big Data and Extreme Scale Computing

The third Big Data and Extreme Scale Computing (BDEC) workshop was held in Barcelona, Spain on January 29th and 30th at the Barceló Sants hotel. This workshop, the third in a series sponsored by the NSF, is premised on the idea that we must begin to systematically map out and account for the ways in which the major issues associated with Big Data intersect with, impinge upon, and potentially change the national (and international) plans that are now being laid for achieving Exascale computing.

The ICL team played a prominent role in this meeting, with Jack Dongarra acting as co-chair, David Rogers designing the meeting’s website and printed materials, and Terry Moore, Tracy Rafferty, and Sam Crawford providing administrative support on site. Members of the Barcelona Supercomputing Center were critical to the meeting’s success and coordinated much of the logistics on the ground; the BSC was also gracious enough to give a tour of their MareNostrum supercomputer, which resides inside a chapel on the BSC campus.

This third BDEC meeting was attended by application leaders confronting diverse big data challenges, alongside members from industry, academia, and government, with expertise in algorithms, computer system architecture, operating systems, workflow middleware, compilers, libraries, languages, and applications.

Barcelona provided an ideal backdrop for the workshop, which was a great success overall, with 40+ individual talks and panel discussions and over 100 attendees from all over the world. Plans are in the works for a comprehensive BDEC report and roadmap, and the foundation for the next iteration of IESP/BDEC meetings is currently being laid.

Recent Releases

2015 ICL Annual Report

icl-report-2015

For fourteen years, ICL has produced an annual report to provide a concise profile of our research, including information about the people and external organizations who make it all happen. Please download a copy and check it out.

Interview

Amina Guermouche Then

Amina Guermouche

Where are you from, originally?

I am originally from Algeria; more precisely I am from the capital, Algiers.

Can you summarize your educational background?

I graduated with a degree in computer science from the university USTHB of Algiers in 2007. I then moved to France to earn a PhD. I started with a Master’s degree at the University of Paris Sud. Then I started a PhD in 2008 which I defended in 2011. I was also a teaching assistant during my PhD. After my defense, I started a Post-Doc at ENS Lyon, where I spent a year. Finally I started another Post-Doc at the University of Versailles that I left last December before moving here.

Tell us how you first learned about ICL.

I had worked work with some members of the team in the past, including George Bosilca, Aurélien Bouteiller, and Thomas Hérault.

What made you want to work for ICL?

I really liked working with George, Aurélien, and Thomas, and after hearing them describe the lab and people here—and learning more about the research objectives—I knew I wanted to join the DisCo team at ICL.

What are you working on while at ICL?

I’m currently working on studying communications in MPI applications in order to fully understand their behavior. After that, I’ll mostly be working on fault tolerance for HPC systems.

If you weren’t working at ICL, where would you like to be working and why?

I would probably have stayed in France and continued my Post-Doc and pursued teaching at the University of Versailles.

What are your interests/hobbies outside work?

I like scuba diving, swimming, watching movies and TV shows, reading, traveling, and hiking.

Tell us something about yourself that might surprise people.

I learned to ride a bicycle just 4 months ago. I joined a biking school in Paris and after two sessions, which last 2.5 hours each, I was able to ride a bike.

Recent Papers

  1. Bouteiller, A., T. Herault, G. Bosilca, P. Du, and J. Dongarra, Algorithm-based Fault Tolerance for Dense Matrix Factorizations, Multiple Failures, and Accuracy,” ACM Transactions on Parallel Computing, vol. 1, issue 2, no. 10, pp. 10:1-10:28, January 2015. DOI: 10.1145/2686892  (1.14 MB)
  2. Bosilca, G., A. Bouteiller, T. Herault, Y. Robert, and J. Dongarra, Composing Resilience Techniques: ABFT, Periodic, and Incremental Checkpointing,” International Journal of Networking and Computing, vol. 5, no. 1, pp. 2-15, January 2015.  (755.54 KB)
  3. Haidar, A., J. Dongarra, K. Kabir, M. Gates, P. Luszczek, S. Tomov, and Y. Jia, HPC Programming on Intel Many-Integrated-Core Hardware with MAGMA Port to Xeon Phi,” Scientific Programming, vol. 23, issue 1, January 2015. DOI: 10.3233/SPR-140404  (553.94 KB)
  4. Dongarra, J., M. A. Heroux, and P. Luszczek, HPCG Benchmark: a New Metric for Ranking High Performance Computing Systems,” University of Tennessee Computer Science Technical Report , no. ut-eecs-15-736: University of Tennessee, January 2015.
  5. Aupy, G., and Y. Robert, Scheduling for fault-tolerance: an introduction,” Innovative Computing Laboratory Technical Report, no. ICL-UT-15-02: University of Tennessee, January 2015.  (416.37 KB)
  6. Haidar, A., T. Dong, P. Luszczek, S. Tomov, and J. Dongarra, Batched matrix computations on hardware accelerators based on GPUs,” International Journal of High Performance Computing Applications, February 2015. DOI: 10.1177/1094342014567546  (2.16 MB)
  7. Anzt, H., S. Tomov, and J. Dongarra, Energy Efficiency and Performance Frontiers for Sparse Computations on GPU Supercomputers,” Sixth International Workshop on Programming Models and Applications for Multicores and Manycores (PMAM '15), San Francisco, CA, ACM, February 2015. DOI: 10.1145/2712386.2712387  (2.29 MB)
  8. Haidar, A., T. Dong, P. Luszczek, S. Tomov, and J. Dongarra, Optimization for Performance and Energy for Batched Matrix Computations on GPUs,” 8th Workshop on General Purpose Processing Using GPUs (GPGPU 8), San Francisco, CA, ACM, February 2015. DOI: 10.1145/2716282.2716288  (699.5 KB)
  9. Haidar, A., P. Luszczek, S. Tomov, and J. Dongarra, Towards Batched Linear Solvers on Accelerated Hardware Platforms,” 8th Workshop on General Purpose Processing Using GPUs (GPGPU 8) co-located with PPOPP 2015, San Francisco, CA, ACM, February 2015.  (403.74 KB)

Recent Conferences

  1. JAN
    George Bosilca
    George
    George Bosilca
  2. JAN
    BDEC Barcelona, Spain
    Jack Dongarra
    Jack
    Sam Crawford
    Sam
    Terry Moore
    Terry
    Tracy Rafferty
    Tracy
    Jack Dongarra, Sam Crawford, Terry Moore, Tracy Rafferty
  3. FEB
    PPoPP 2015 San Francisco, California
    Hartwig Anzt
    Hartwig
    Piotr Luszczek
    Piotr
    Hartwig Anzt, Piotr Luszczek
  4. FEB
    SI^2 PI Meeting Arlington, Virginia
    George Bosilca
    George
    George Bosilca

Upcoming Conferences

  1. MAR
    MPI Forum Portland, Oregon
    Aurelien Bouteiller
    Aurelien
    Aurelien Bouteiller
  2. MAR
    SUPER all hands meeting San Diego, California
    Anthony Danalis
    Anthony
    Anthony Danalis
  3. MAR
    SIAM CSE Salt Lake City, Utah
    Azzam Haidar
    Azzam
    George Bosilca
    George
    Hartwig Anzt
    Hartwig
    Ichitaro Yamazaki
    Ichitaro
    Piotr Luszczek
    Piotr
    Stanimire Tomov
    Stan
    Azzam Haidar, George Bosilca, Hartwig Anzt, Ichitaro Yamazaki, Piotr Luszczek, Stanimire Tomov
  4. MAR
    GTC15 San Jose, California
    Azzam Haidar
    Azzam
    Piotr Luszczek
    Piotr
    Stanimire Tomov
    Stan
    Azzam Haidar, Piotr Luszczek, Stanimire Tomov
  5. MAR
    Jack Dongarra
    Jack
    Jack Dongarra

Recent Lunch Talks

  1. JAN
    8
    Tony Hey
    Tony Hey
    The Fourth Paradigm: Data-Intensive Scientific Discovery, Open Science and the Cloud PDF
  2. JAN
    16
    Emmanuel Jeannot
    Emmanuel Jeannot
    INRIA
    Topology Aware Data Management PDF
  3. JAN
    23
    George Bosilca
    George Bosilca
    Building Blocks for Resilient Applications PDF
  4. FEB
    6
    Amina Guermouche
    Amina Guermouche
    FoREST-mn: Runtime DVFS Beyond Communication Slack PDF
  5. FEB
    13
    Yves Robert
    Yves Robert
    Scheduling Computational Workflows on Failure-prone Platforms PDF
  6. FEB
    27
    Piotr Luszczek
    Piotr Luszczek
    Deep Neural Networks for Image Classification – A Primer PDF

Upcoming Lunch Talks

  1. MAR
    6
    Azzam Haidar
    Azzam Haidar
    Performance Bounds in Symmetric Eigenvector Calculations PDF
  2. MAR
    13
    Audris Mockus
    Audris Mockus
    EECS
    Evidence Engineering PDF
  3. MAR
    20
    Anthony Danalis
    Anthony Danalis
    Using PaRSEC to Develop Non-static Applications
  4. MAR
    27
    Yves Robert
    Yves Robert
    INRIA
    Voltage Overscaling Algorithms for Energy-Efficient Workflow Computations With Timing Errors PDF

People

  1. Amina Guermouche
    Amina Guermouche joined ICL at the beginning of 2015 as a research associate, and will be working with the DisCo team. Welcome!
  2. Damien Genet
    Damien Genet joined ICL on February 1st as a new post-doc working for the DisCo team. Welcome, Damien!
  3. Yaohung Tsai
    Yaohung "Mike" Tsai joined ICL at the beginning of 2015 as a Graduate Research Assistant to work with the Linear Algebra group. Welcome aboard, Mike!

congratulations

Shumin Jia

Shumin Jia was born to Yulu and Rui Jia on December 22, 2014. Shumin is 20.25 inches in length and weighs 7 pounds, 5 ounces. Both mother and baby are doing well. Congratulations to the Jia family!

shumin_jia