Computing experts converge on East Tennessee to talk computing, AI at SMC24

155 attendees from all over the world gathered for SMC24 for a wide range of presentations from industry leading experts. Credit: ORNL, U.S. Dept. of Energy

The Smoky Mountain Computational Sciences and Engineering Conference, or SMC24, entered its third decade with the 21^st annual gathering in East Tennessee.

Hosted by the Department of Energy's Oak Ridge National Laboratory, conference attendees gathered in downtown Knoxville to hear from computing leaders from the private sector, government agencies and academia who traveled from across the country and internationally to be part of the three-day event.

The conference boasted 155 attendees from 41 institutions, with this year's theme: "Accelerating science and engineering discoveries through advanced technologies and integrated research infrastructures for experiments, machine learning/ artificial intelligence, and modeling and simulation."

SMC24 attendees enjoyed cutting edge insights into the latest advancements in computer science and engineering. Credit: ORNL, U.S. Dept. of Energy

"SMC has come of age as a go-to conference," said Mallikarjun Shankar, who serves as division director for the National Center for Computational Science at ORNL. "It was a dynamic meeting including an AI ethics data-challenge, integrated research demos, provocative panels and a terrific line up of cutting-edge science and technology talks. As always, it was wonderful to have this community come together to discuss the science and technology challenges and opportunities of the day and debate over future directions."

Mirroring the broader computing landscape, AI was the dominant discussion point for the conference that covered a wide range of topics, including the vast utility of large language models, AI's relationship with high performance computing and maintaining security and privacy around data.

"SMC24 upheld its reputation for excellence, delivering groundbreaking insights into accelerating science and engineering discoveries through advanced technologies and integrated research infrastructures," said conference organizer and ORNL senior computer scientist Oscar Hernandez. "The event encouraged collaboration across participants through its talks, panel discussions and social gatherings while exploring the future of scientific computing through emerging technologies like AI and quantum computing to solve critical challenges in a safe, trustworthy and energy-efficient way."

Highlighting Day 1 were presentations from Jacques Pienaar, ML Systems Research team lead at Google Deepmind, and Prasanna Balaprakash, director of AI programs at ORNL. Balaprakash walked attendees through ORNL's rich history of leading the way in advancing artificial intelligence.

The Smoky Mountain Mountains Computational Sciences and Engineering Conference is one of the leading events for exploring infrastructures for cutting edge experiments, artificial intelligence, and modeling and simulation. Credit: ORNL, U.S. Dept. of Energy

"SMC serves as a critical nexus, bringing together the laboratory, academic and industry communities to explore cutting-edge computational science and emerging topics such as AI," Balaprakash said. "It remains an exceptional platform for exchanging ideas and fostering new collaborations in these rapidly evolving fields."

Day 1 also featured the annual student "Data Challenge." The 2024 edition was a change from previous years where contestants chose from a series of data challenges to solve. Instead, finalists submitted essays as part of the Trustworthy AI for Science Essay contest and detailed their findings in a Ted Talk-style presentation. This year's finalists included three students from nearby Farragut High School.

"I think the format change made the contest more accessible to all career levels since we had finalists from high school, college, graduate school and senior career research represented among our finalists," said Suzanne Parete-Koon, a high-performance computing engineer at ORNL who led the data challenge contest. "We hope this essay contest continues to grow and capture ideas about trustworthy, reproducible, ethical and energy efficient AI from a broad set of perspectives."

Panelists at SMC24 took questions and debated some of the most pressing topics in computational science and engineering today. Credit: ORNL, U.S. Dept. of Energy

Days 2 and 3 of SMC24 included several presentations and panel discussions from ORNL and other national laboratory researchers, along with representatives from major computing organizations and universities.

For ORNL Director of Science-Security Initiative Integration Shaun Gleason, the learning opportunities are only part of what makes SMC24 such a special event.

"This is my twelfth time attending SMC, and it's always been a very productive and educational event," Gleason said. "One of the best things about SMC24 was networking with professional colleagues across the DOE lab system (including my own ORNL staff that I don't get to see on campus!), government sponsors, university partners and industry representatives that are developing leading edge computing and software."

Networking opportunities extended beyond the conference hours, beginning with a dinner at the iconic Sunsphere after Day 1, followed by a relaxed gathering at the rooftop Five Thirty Lounge to wrap up Day 2. Both establishments are in Knoxville's historic downtown.

SMC24 sponsors include Advanced Micro Devices (AMD), Dell Technologies, Hewlett Packard Enterprise, Intel, IBM, Microsoft, Amazon Web Services (AWS) and NVIDIA.

UT-Battelle manages ORNL for the Department of Energy's Office of Science, the single largest supporter of basic research in the physical sciences in the United States. The Office of Science is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science. - Mark Alewine