Argonne Leadership Computing Facility

Today we’re excited to shine a spotlight on Aaqel Shaik for our summer student series. Aaqel is a rising senior studying Computer Science at University of Illinois Chicago. This summer, he helped build an AI-powered personal assistant that helps scientists conduct research by facilitating various day-to-day tasks like simulation, summarization and general information retrieval. He worked with cutting-edge #XR technology including the Apple Vision Pro. “There were many times during the last few weeks where I had an idea for what I wanted to implement in my head, but it wasn't immediately obvious how I should go about making it. One such moment was when I wanted to build a real-time data-streaming pipeline between the XR UI and the backend simulation, so users could see changes in the simulation as they happened. This led me to discover producer-subscription models, and I believe it is one of the most useful things I've learnt at my time here in Argonne National Laboratory.” “XR is one of those exciting fields where, unfortunately, the cost of entry is higher than most other fields in Computer Science. I wasn't really considering XR development as a possible career path, but I've had so much fun these past few weeks and I'm interested in exploring it further in my own time.” Learn more about student opportunities at the ALCF: https://lnkd.in/ggN2E4Vv

Learn how Argonne National Laboratory researchers are using AI and exascale computing power to advance connectomics research. While mapping the human brain will take many more years of work, the team will leverage the ALCF’s Aurora supercomputer to develop methods for mapping smaller brain samples, laying the foundation for future explorations - https://lnkd.in/dczcg3jP "(The brain) is one of the most complex things on the planet. It’s certainly the most complex thing in our bodies, and we don’t totally understand how it works. What we’re trying to do is reconstruct its structure and connectivity." - ALCF’s Thomas Uram

Meet Sreten Kljaic, an ALCF summer student studying Computer Science at University of Illinois Chicago. This summer, Sreten created a Python interface for a high energy physics collider simulation engine. This interface incorporates machine learning into the simulator to make it more user friendly. We caught up with Sreten to ask about his summer work. He told us that one “aha” moment he had was gaining a new perspective on large code bases: “I haven't worked with larger code bases before so I wasn't familiar with build tools. Once I became more comfortable with #CMake it made everything straightforward and allowed me to tackle relatively complex dependencies effortlessly.” Another breakthrough for Sreten was understanding “the importance of being a good communicator and that all projects depend on collaboration” Learn more about student opportunities at the ALCF: https://lnkd.in/ggN2E4Vv

Join us on September 25, 2024, for a webinar covering Intel Software's performance profiling tools on Aurora. Aurora is one of the most powerful systems in the world, and the Intel performance profiling tools are essential for maximizing the system's throughput with applications for science and engineering. In this webinar, ALCF's JaeHyuk Kwack will provide an overview of the tools and instructions on utilizing them with Aurora's applications. https://lnkd.in/gm9vNFMW

Today we’re shining a spotlight on Fredric Massa (Ricky), a rising senior from University of Illinois Chicago. As an ALCF summer student, Ricky helped develop an AI personal assistant on the Apple Vision Pro specifically tailored for scientists engaged in experimentation and research. In this project, Ricky got hands on experience working with XR technology and the supercomputer Polaris. A key learning moment this summer for Ricky was “Figuring out how to submit jobs to Polaris from our Apple Vision Pro app and show the results in the app. When I began working on this, I was unsure where to start. I began by looking into Swift packages, but found nothing that worked with VisionOS. I ended up setting up a server that handles everything concerning Polaris, that our app sends HTTP requests to.” Learn more about student opportunities at the ALCF: https://lnkd.in/ggN2E4Vv

Applications are being accepted for the 2025 BSSw Fellowship!! Get recognized for your passion around high-quality scientific software to improve scientific productivity. Receive up to $25,000 for an activity that promotes better #scientificsoftware. Activities can include organizing a workshop, preparing a tutorial, or creating content to engage the scientific software community. https://lnkd.in/gs7wbzXd Lawrence Livermore National Laboratory Berkeley Lab National Energy Research Scientific Computing Center (NERSC) Sandia National Laboratories Argonne National Laboratory Oak Ridge National Laboratory Los Alamos National Laboratory U.S. Department of Energy (DOE) National Science Foundation (NSF) Mary Ann Leung #somycodewillseethefuture #bssw #rse #DOE #NSF

For today’s summer student spotlight, Amy Byrnes, a Computer Science graduate student from University of Illinois at Chicago, shared some of her favorite moments at the ALCF with us. She spent her summer exploring different tools for writing parallel code and compared strategies for parallelizing numerical methods for solving partial differential equations. She also began exploring the connections between HEP and #HPC, in preparation for her upcoming PhD work. “I've been able to run my code on Polaris, giving me first-hand experience not only with the power of a world-class HPC system, but also the logistics of getting a job running on one. I've also been able to have make connections with mentors and peers that I look forward to maintaining for years to come.” “I've had a lot of little moments where an idea I had for how my code should work went from being hazy to fully formed -- that tended to happen when I gave a potential solution time to ‘marinate’ rather than trying to force it to come together immediately." “But my biggest ‘aha’ moment came during a regular update meeting. I had been really struggling with figuring out what direction I wanted to go with my research beyond my time at Argonne National Laboratory. A supervisor made a comment to another student about a certain topic, and though that topic had been mentioned before, something about the way it was phrased that time just rang a bell in my head -- *this* was the thing I wanted to dig into! I think it just goes to show that progress is often hard to predict.”

Meet Xiongxiao Xu, an ALCF summer student from Illinois Institute of Technology. Xiongxiao joins us as a Ph.D. candidate studying Computer Science. This summer, he focused on performance-energy trade-offs in heterogenous #HPC systems: “CPU/GPUs in supercomputers consume a great deal of energy at maximum frequency. Decreasing frequency can decrease power consumption, but not necessarily energy consumption, because the execution time will be extended at lower frequency. That is why we call it performance-energy trade-off. Also, the energy consumption optimization in HPC is very significant." Xiongxiao says he enjoyed modeling real-world data from supercomputers, something he would not have had the chance to do in a traditional academic setting. “The experience from Argonne National Laboratory let me know the energetic and healthy work environments at national labs. I also have a chance to talk and connect to world-renowned scientists. It makes me have a deeper understanding in research and inspire my interests of working in academia.” Learn more about student opportunities at the ALCF: https://lnkd.in/ggN2E4Vv

Today we’re excited to shine a spotlight on our summer student Revathi Dhotre, a master’s student studying Computer Science at University of Illinois Chicago. This summer, she analyzed #supercomputer logs to uncover patterns and anomalies, ensuring that they ran smoothly and pushing the boundaries of what's possible in high-performance computing. To make sense of the data, Revathi leveraged machine learning algorithms that acted as sophisticated pattern-recognition tools. By combining the power of machine learning and clustering techniques, she was able to transform raw log data into actionable insights. “Working at the Argonne Leadership Computing Facility offered invaluable insights into my career and the field of high-performance computing. One of the most striking revelations was the scale at which #HPC operates. Handling massive datasets and complex systems underscored the importance of collaboration with a diverse team. It became clear that solving sophisticated problems requires not just individual expertise but also collective efforts from data scientists, engineers, and researchers working together.” Learn more about student opportunities at the ALCF: https://lnkd.in/ggN2E4Vv

Meet Izzet Yildirim, a Computer Science Ph.D. Candidate from Illinois Institute of Technology. As a summer student at the ALCF, Izzet worked on validating and improving DLIO, a benchmark tool for simulating the input/output behavior of scientific #deeplearning applications. His work involved running deep learning workloads, analyzing their behavior, and comparing it to DLIO simulations to identify areas for improvement in the benchmark's accuracy and capabilities. “Working at Argonne has provided me with firsthand experience of the complexities involved in large-scale scientific computing. I have seen how challenges at this level are far more intricate than those encountered on smaller scales. Overcoming these challenges has been an eye-opening experience, significantly broadening my understanding and approach to problem-solving. My work on the Polaris system this summer allowed me to deeply understand the critical role of high-performance computing in tackling large-scale scientific problems. This experience has made me eager to see how Aurora will further enhance scientific discovery, and I look forward to the opportunity to work with it in the future.” Learn more about student opportunities at the ALCF: https://lnkd.in/ggN2E4Vv