Name: Kevin Brown
Pronouns: he/him/his
Biography:
Kevin A. Brown is an Argonne Scholar – Walter Massey Fellow in the Mathematics and Computer Science Division at Argonne National Laboratory. Kevin received his Ph.D. from the Tokyo Institute of Technology. At Argonne, Kevin evaluates new networking technologies and designs for next-generation supercomputers as well as improves network simulators used for evaluating supercomputer designs. In the area of evaluation, he works on creating new performance analysis tools and techniques to measure and analyze the performance of AI and climate prediction applications on large supercomputers, focusing mainly on network and I/O traffic activities. In the area of network simulation, he simulates potential designs for future supercomputer to identify promising design choice while also creating new simulation techniques to improve the efficiency of networks simulations. His current day-to-day activities involve coding new performance measurement tools, creating simulation models of workloads and system hardware, combining parallel discrete event simulations (PDES) with surrogate models, and evaluating and optimizing data movement (network and I/O) strategies.
Institution/Lab: Argonne National Laboratory
Website: https://www.anl.gov/profile/kevin-a-brown
SRP Collaboration Topic/Title: Investigating and Optimizing Supercomputer Network Performance
Field or research area: Computer Science
Please select all the topical areas that apply to your project:
Computer Science (i.e., architectures, compilers/languages, networks, workflow/edge, experiment automation, containers, neuromorphic computing, programming models, operating systems, sustainable software); Data Science (i.e., data analytics, data management & storage systems, visualization); High-Performance Computing; Machine Learning and AI
Brief Abstract:
Supercomputers use the fastest network and storage technologies to help solve big problems in artificial intelligence (AI) and climate science among other fields. However, properly configuring supercomputers requires carefully tuning the hardware and software to meet the performance needs of the work to be done. Projects ideas include creating new software to measure performance on supercomputers, analyzing and optimizing the performance of climate science and AI applications, investigating and simulating new supercomputer designs, evaluating workflows between remote scientific instruments and supercomputers, and improving discrete event simulations.
Desired relevant skills, background, or interests:
Specific background depends on the specific project focus. Generally, experience with C or Python may help along with an interest in problem solving
Other comments:
This is a placeholder for a range of project ideas in the area of performance analysis, network simulations, and supercomputer designs. We can come up with specific projects based on your skills and interests.
Do any special requirements apply? Permanent Resident OK; International OK
Other, specify:
Keywords:
Supercomputer; AI; Networking: Performance analysis;
Lightning Talk Title: Evaluating and Simulating Supercomputers … other scientific infrastructure