News & Events

Share          

​Hitting the Mark at the Student Cluster Competition 2018

Published on: 11-Jul-2018

From left to right: The SCSE team comprises of Zhang Xinye (second from the left), Tan Ying Hao (Vice-Captain), Heng Weiliang, Shi Ziji (Captain), Wu Bian and Hao Meiru.​​​

Students from the School of Computer Science and Engineering (SCSE), Nanyang Technological University, hit the mark again at the recent Student Cluster Competition held in conjunction with the International Supercomputing Conference 2018 (ISC'18) at Frankfurt, Germany. 

The competition attracted a total of 12 teams from 9 countries such as South Africa, China, Poland, UK, etc. Despite the intense competition, the SCSE team distinguished themselves by taking the 2nd highest overall score.

Mentored by Assoc Prof Lee Bu-Sung, Francis, the SCSE team consists of 6 undergraduate students namely, Shi Ziji (Captain), Tan Ying Hao (Vice-Captain), Wu Bian, Zhang Xinye, Heng Weiliang and Hao Meiru. During the competition, teams are required to optimise and run a number of applications from different domains such as fluid dynamics, quantum physics, and computer vision. 

Ziji, captain of the team, said that it was a hard-won award. They met regularly in weekly discussions and tuned the application and system. To cater to various application requirements, the team took a different hardware configuration, favouring a heterogeneous cluster unlike majority of the other teams which feature homogenous clusters. The SCSE team's configuration has the least number of CPUs and GPUs within the top-scoring teams, yet it worked out well in the competition.

Ying Hao, vice-captain of the team who oversaw the Deep Learning application, commented that it was very challenging to tune the TensorFlow software to maximise the speed of learning on the ImageNet dataset using the NVIDIA GPU cards while ensuring that the cluster stays within the power limit of 3KW, a constraint set by the organisers. With only 10 GPU cards, the SCSE team achieved a stunning result of 3170 images per second on the VGG16 model, beating many teams with 16 GPU cards.

The team's participation in the competition was made possible through the sponsorship from Dell EMC, Nvidia Technological Center Singapore, National Supercomputing Centre Singapore, Mellanox Technology and JOS. The team would also like to express gratitude to Assoc. Prof. Francis Lee for mentoring the team, Ms. Irene Goh from the Parallel and Distributed Computing Lab and Asst. Prof. Lin Guosheng for his guidance on the Deep learning challenge. 

Back to listing