OVERVIEW

Academia Sinica Grid Computing Centre (ASGC) was established in 2005, whose founding goal is to build up the global distributed computing infrastructure for Large Hadron Collider (LHC) experiments – Worldwide LHC Computing Grid (WLCG), coordinated by CERN. ASGC aims to support the fast-growing needs of scientific computing and big data analysis in Academia Sinica by the advanced distributed computing technologies learned from WLCG. On the other hand, serving as the WLCG Tier-1 Center in Asia, ASGC has been working with WLCG sites in Japan, Korean, China, Hong Kong, Malaysia, Thailand, India, Australia, New Zealand, Pakistan and Taiwan to enhance the capacity and efficiency of the global collaborations persistently.

ASGC has been the primary scientific computing facility in Academia Sinica since its foundation. ASGC works closely with research groups to integrate the analysis pipeline, data management lifecycle and analysis software etc. with the campus distributed infrastructure. Both application and system efficiency are improved progressively according to the user experiences and evolved ICT technologies. Started from ATLAS and CMS experiments collaborations in the beginning, ASGC has been supporting around 30 research groups of multiple disciplines in Academia Sinica in 2021. Statistically, the compound annual growth rate of CPU usage at ASGC between 2012 and 2021 is about 15%. The GPU computing applications has very sharp growth rate by over 20 times CAGR during 2018 and 2021.

Reliability and efficiency are the key objectives of ASGC. Driving by the requirements of user communities in various domains, ASGC keeps advancing the application efficiency systematically by the evolved ICT technologies such as GPU and container computing, machine learning and big data analysis over distributed cloud learned from international collaborations from WLCG and EGI. Improvement in research data management as well as the exploration towards open science ecosystem are also included in ASGC milestones.

▶ Acting as a WLCG Tier-1 Center and jointly developing advanced distributed computing infrastructure and technologies for O(1000)PB big data analysis globally through the collaborations with WLCG.

▶ Enhancing science-enabling capability by advanced big data analysis capability over distributed cloud.

▶ Achieving system efficiency optimization intelligently, covering power, thermal, application and ICT systems