Dec. 1, 2022 — A team of students from the San Diego Supercomputer Center (SDSC) and the Computer Science & Engineering (CSE) Department at UC San Diego has won the 2022 Student Cluster Competition HPL Benchmark Contest, placing third overall in the competition at the recent 2022 Supercomputing Conference in Dallas, Texas.
Six UC San Diego undergraduate students – Matthew Mikhailov (team lead), Stefanie Dao, Anish Govinda, Michael Granado, Yuchen Jing and Davit Margarian, as well as two alternates: Rachel Hadran and Khai Vu – made up the team. Named 2MuchCache, the team was mentored by Mary Thomas (SDSC) and Bryan Chin (CSE), with assistance from Andreas Goetz, Mahidhar Tatineni and Bob Sinkovits.
“The competition consisted of the team running four HPC benchmarks and four science applications. The High-Performance Linpack benchmark implementation requires that each competitor runs a well-known code to solve a uniformly random system of linear equations on their customized cluster and to then report time and floating-point execution rate using a standard formula for operation count,” explained Thomas, SDSC’s HPC training lead. “The HPL Benchmark Contest at SC22 included 13 talented worldwide teams, so we were very pleased to win first place.”
The team won the HPL Benchmark competition by reaching 114.26 Teraflops under a 3000-watt power limit, running on a SuperMicro system consisting of four AMD Insticnt GPUs and 128 cores of AMD EPYC 7773X CPUs.
To learn more about 2MuchCache, their SC22 poster shares details, and more information can be found on the team website.
The Student Cluster Competition was developed in 2017 to immerse undergraduate and high school students in high performance computing (HPC). SCC teams consist of worldwide participants, in a non-stop, 48-hour challenge to complete a real-world scientific workload, while keeping the cluster up and running, and demonstrating to the judges their HPC skills and knowledge. Each SCC team consists of six students who design and build a small cluster with support from mentors, as well as hardware and software industry partners. The teams learn designated scientific applications and apply optimization techniques for their chosen architectures.
To learn more about the SCC series, see the Student Cluster Competition website.
Source: Kimberly Mann Bruch, SDSC