Guided Interest Groups (GIGs) are community learning experiences where SCALE participants shepherd interested students through the SC technical program.
- GIGs are open to all students attending the conference; sign-ups will open first for students participating in the Students@SC Cohorts.
- Students can register for a GIG which is lead by a pair of SCalers.
- Participating students are expected to attend sessions planned by the GIG leaders as well as pre- and post-session discussions facilitated by the session lead.
- Once a student has registered and received a registration confirmation for a specific GIG, they will be put in contact with the SCalers leading their GIG. The SCalers will send all meeting invites to the GIG participants.
Please take the time to review the different GIGs and submit your selection.
GIG 1 (Virtual)
From Linear Algebra to Distributed Deep Learning: How Do Machines Learn?
Leads
Ana Solórzano & Pratik Nayak
Ana Solórzano
Federal University of Rio Grande do Sul, Brazil Computer Science Department
I am a Master’s student in Computer Science at UFRGS. I became interested in High-Performance Computing during my Bachelor’s in Computer Science, when I did an internship to parallelize an algorithm in Astronomy for hybrid environments. I have been researching in projects on performance analysis of scientific applications and distributed deep learning frameworks. Besides working in HPC, I advocate for more diversity in tech, co-founding Brazilians in Tech. I love reading, watching movies and going out with my friends.
Pratik Nayak
Karlsruhe Institute of Technology, Germany
I am a doctoral candidate at the Karlsruhe Institute of Technology, Germany working on Numerical Linear Algebra and High Performance Computing. I am particularly interested in developing algorithms and software for HPC applications. Currently, I am involved in the Exascale Computing Project, work on developing Ginkgo, a HPC library and work on developing asynchronous iterative methods for solving linear systems. I enjoy running, hiking, playing tennis and watching sitcoms.
Description
Have you ever wondered how machines learn? How do the algorithms that you hear of in autonomous driving, face recognition, and protein folding function? In this GIG, we give you a taste of some of the interesting aspects of Machine Learning and Linear Algebra through SC21.
You will get to check out the state-of-the-art of Deep Learning tools and how they are using HPC power to deliver faster and more accurate solutions. You will also get inspired by the annual SC keynote and the two talks by AI researchers working in the industry.
Schedule (Central Standard Time)
MON, NOV 15
- 8:30 am CST: Introduction and Daily Kick-off Meeting
- 9 am CST: Keynote: Exascale Computing Will Change the World (ML in HPC Workshop)
- 12 pm CST: 7th Workshop on Machine Learning in High Performance Environment: High-Performance Deep Learning Toolbox for Genome-Scale Prediction of Protein Structure and Function (Fourth presentation)
- 12:30 pm: End-of Day-Meeting
TUE, NOV 16
- 8 am CST: Daily Kick-off
- 8:30 am CST: SC21 Keynote: Computing and the Humanities
- 10:30 am CST: Papers Session: Efficient Deep Learning Tools (Optional)
- 3:30 pm CST: Papers Session: Hardware Efficient Deep Learning (Optional)
- 5 pm CST: End-of-Day Meeting
THU, NOV 18 (Choose one of the two 10:30 am talks)
- 10 am CST: Daily Kick-off
- 10:30 am CST: Invited Talk: 1-AI4SCinece: Convergence Between AI and Scientific Computing
- 10:30 am CST: Invited Talk: The Growth of AI: Implications on Hardware, Systems, Software, and the Environment
- 1:30 pm CST: Paper: CAKE: Matrix Multiplication Using Constant-Bandwidth Blocks
- 2 pm CST: End-of-Day Meeting
FRI, NOV 19
- 10 am CST: Daily Kick-off
- 10:30 CST: Panel: Reproducibility in HPC: Passing Fad or a Work in Progress?
- 12 pm CST: End-of-Day Meeting
GIG 2 (Virtual & In-Person)
HPC at a Glance
Leads
Kae Suarez, Jonas Posner, & Kunj Champaneri
Kae Suarez
University of Tennessee, Knoxville
I am a PhD student advised by Dr. Michela Taufer, researching in the GCLab at the University of Tennessee, Knoxville. I got my B.S. in Computer Science in 2020 at UTK. My passion for HPC most lies within areas of reproducibility, correctness, and usability. This leads me to working with non-determinism, HPC migration for domain scientists, and creation of software infrastructure for domain scientists. Outside of HPC, I have a passion for LGBT+ issues, and have worked to help with efforts in that field where I can. For fun, I have interests in TTRPGs, speed running, and games in general, and hope to pick up some board games once I’m seeing more people in person!
Jonas Posner
University of Kassel, Germany
I am a last-year PhD student working at the University of Kassel, Germany. My research centers around HPC, with focus on PGAS, task-based parallel systems, load balancing, resilience, and resource elasticity. I have participated in SC as a volunteer for the past four years and am very much looking forward to SC21, as it is always a lot of fun to meet friends from previous years, meet new interesting people, and of course learn many new things about science!
Kunj Champaneri
Slippery Rock University, Pittsburgh, PA
I am currently doing my bachelor’s in Computer Science from Slippery Rock University of Pennsylvania. This is my 3rd SC experience. I was a regular student volunteer in SC19 and SC20. I have been working with high-performance computing since my 2nd year of bachelor’s and exploring topics like scheduling, and algorithms for optimization. In addition, I am also interested in computer vision and cybersecurity. I like to solve puzzles like Rubik’s Cube and Megamix.
Description
This GIG will be an overview of the state and applications of High Performance Computing (HPC). It is intended to give participants an at a glance view at the importance of the field of HPC, along with directions to consider as they move into the field. Applications of HPC explored in this GIG include general COVID19 research and molecular dynamics (MD) simulation, and future directions include sustainability and artificial intelligence (AI). The goal is to give a high-level glance at the field and its trajectory to encourage curiosity and entry into the field of HPC instead of focusing on specific implementations or algorithms. Discussions will center around answering introductory questions and how to approach HPC as a prospective researcher.
Schedule (Central Standard Time)
SUN, NOV 14 | Kick-off Meeting
- 10 am CST
TUE, NOV 16 | Panel: HPC’s Growing Sustainability Challenges and Emerging Approaches
- 1:15 pm CST: Pre-session Meeting
- 1:30 pm CST: Panel
- 3 pm CST: Post-session Meeting
TUE, NOV 16 | Gordon Bell Finalist: Anton 3: Twenty Microseconds of Molecular Dynamics Simulation Before Lunch
- 3:15 pm CST: Pre-session Meeting
- 3:30 pm CST: Talk
- 4 pm CST: Post-session Meeting
WED, NOV 17 | Invited Talks: 1) High Performance Convergence Computing, and 2) HiPEAC: 17 Years of Growing an HPC Community in Europe
- 10:15 am CST: Pre-session Meeting
- 10:30 am CST: Invited Talk 1
- 11:15 am CST: Invited Talk 2
- 2 pm CST: Post-session Meeting
THU, NOV 18 | Invited Talks: 1) The Growth of AI: Implications on Hardware, Systems, Software and the Environment, and 2) The Rise of Supernetwork Data Intensive Computing
- 10:15 am CST: Pre-session Meeting
- 10:30 am CST: Invited Talk 1
- 11:15 am CST: Invited Talk 2
- 12 pm CST: Post-session Meeting
FRI, NOV 19 | Panel: HPC Responses to the COVID-19 Pandemic
- 8:15 am CST: Pre-session Meeting
- 8:30 am CST: Talk
- 10 am CST: Post-session Meeting
GIG 3 (In-Person)
Enabling Scientific Discovery with HPC
Leads
Nigel Tan & Daniel Barry
Nigel Tan
University of Tennessee, Knoxville
I am a PhD student at UTK and a member of Global Computing Lab under Dr. Michela Taufer. I received my B.S. from UC Merced and my masters from Rice University. My primary interests are in performance portability for scientific applications. My research aims focuses on high performance simulations and checkpointing on heterogeneous systems. Additionally I also study techniques for addressing non-determinism in HPC. I enjoy hiking and games. I am looking forward to reconnecting with colleagues and helping students join the HPC community.
Daniel Barry
University of Tennessee, Knoxville
I am a PhD student in Data Science and Engineering. I do research with the Innovative Computing Lab (ICL) in the Performance Tools group. I received my Bachelors of Science in Computer Engineering at the University of Tennessee. My research interests include application performance monitoring, numerical methods, and large-scale data analytics. I competed in the Student Cluster Competition in 2013 and have been part of the HPC community ever since. I am an avid fan of food and cooking!
Description
This GIG aims to explore how high performance computing (HPC) is leveraged for making scientific advances. HPC is a vital tool for researchers. Modeling and simulation drive discovery where experimentation is too costly. Physical experiments produce vast amounts of data that often require HPC scale resources to process and analyze. Students will explore how scientists use HPC in various scientific domains to accelerate and enhance their research. Experienced mentors will guide students through several conference events.
Schedule (Central Standard Time)
SUN, NOV 14 | CAFCW21: Computational Approaches for Cancer Workshop 2021
- 8:30 am CST: Kick-off Meeting
- 9 am CST: Workshop
- 10:45 am CST: Post-session Meeting
TUE, NOV 16 | SC21 Keynote: Computing and the Humanities
- 8 am CST: Pre-session Meeting
- 8:30 am CST: Keynote
TUE, NOV 16 | Paper Session: Computational Biology
- 10 am CST: Pre-session Meeting
- 10:30 am CST: Paper Session
- 12 pm CST: Post-session Meeting
TUE, NOV 16 | Invited Talk: Going Slow to Go Fast: An Integrated Approach to COVID-19 Therapeutic Discovery at the DOE National Laboratories
- 1 pm CST: Pre-session Meeting
- 1:30 pm CST: Invited Talk
- 2:15 pm CST: Post-session Meeting
WED, NOV 17 | Gordon Bell Prize Finalists Session 2
- 3:30 pm CST: Pre-session Meeting
- 4 pm CST: Talks
- 5 pm CST: Post-session Meeting
THU, NOV 18 | Paper Session: Data Compression and Workflows
- 3 pm CST: Pre-session Meeting
- 3:30 pm CST: Paper Session
- 5 pm CST: Post-session Meeting