麻豆影视文化传媒在线看|果冻传媒91制片厂麻豆|性色网站|国产成人吃瓜网|麻豆文化传媒百度云|韩国黄色一级黄色片|成人电影区|糖心vlog是真的吗|黄瓜视频丝瓜视频香蕉视频|国产精品视频在一区鲁鲁,性感丰满美乳巨乳,蜜桔影院91制片厂,爱豆传媒陈可心作品名字

Home>LATEST NEWS

?Tsinghua’s Student Cluster Team claims championship at ISC25

The Tsinghua University Student Cluster Team emerged victorious at the ISC 2025 Student Cluster Competition after securing the overall championship and the Highest LINPACK Benchmark Score award. This marks Tsinghua’s 19th win across the three major international student supercomputing competitions and their 8th championship in the ISC competition since its inception in 2012. The competition was held at the Congress Center Hamburg, Germany.

The ISC25 competition challenged teams to build and optimize computer clusters under a 6-kilowatt power constraint. The tasks included running benchmark programs such as LINPACK, HPCG, and IO500, as well as scientific applications like the earthquake simulation software SeisSol, the post-training of large language model Llama, the computational fluid dynamics software code_saturne, and a molecular dynamics simulation application LAMMPS, which is a secret application announced on site.

Facing technical challenges and jet lag, the Tsinghua team demonstrated exceptional system expertise, teamwork, and adaptability. They strategically configured a 4-node, 12-GPU cluster, achieving a record-breaking 340 TFLOPS (teraflops) in the LINPACK benchmark, securing first place in that category and setting a new competition record.

The ISC25 competition, co-organized by the HPC-AI Advisory Council and ISC Group, stands as one of the world’s top three student supercomputing contests alongside SC and ASC, featuring 10 teams from 10 countries or regions in the in-person finals.

The Tsinghua team consisted of six undergraduate students: Yuqing Yang, Shuangyu Li, Jiaxin Lv from the Department of Computer Science and Technology, and Kai Yang, Chaopeng Jia, and Yang Zhang from Zhili College. An additional three students Zhangqiao Yue, Shi Chengyu, and Yucheng Wang participated in the training. The team was instructed by Wentao Han, Yuyang Jin, and Professor Jidong Zhai, with technical support from doctoral students Runqing Zhang, Jingbo Shan, and Shengqi Chen from the Institute of High-Performance Computing.

Editor: Li Han

Copyright 2001-2021 news.tsinghua.edu.cn. All rights reserved.