9:00 AM - 9:15 AM
[MGI31-01] Research Infrastructure Accelerating Scientific Research Utilizing Large-scale Data
★Invited Papers
Keywords:Supercomputing system, Data Aggregation Infratstructure , High-speed Data Transfer Service
The advancement of measurement technologies has significantly increased the amount of data obtained through scientific measurement devices. Also, computing technologies has been advancing. As the result, recent scientific research has been increasing scientific data to be treated. For example, in order to understand scientific phenomena, scientists needs to observe what happens temporally and spatially at high resolution through computer simulations. Furthermore, the recent success of AI (Artificial Intelligence) techniques characterized with DL (Deep Learning), ML (Machine Learning) and LLM (Large Language Models) has been gathering scientists’ concerns and expectations to application of such techniques to their own large-scale scientific data. For a reason, supercomputing infrastructure which are able to accommodate scientists’ large-scale computing needs and requirements from both HPC (High-Performance Computing) and AI purposes and data infrastructure which are able to satisfy scientists’ data utilization needs and requirements.
From the background above, the D3 Center (former: Cybermedia Center) at the University of Osaka, which have a responsibility to provide supercomputing services to researchers in Japan, not limited to the University of Osaka, has been working on the development and deployment of a next-generation research infrastructure accelerating data-driven scientific research utilizing large-scale data. Through the realization of such a next-generation research infrastructure, we hope that it facilitates scientists and researchers to utilize our supercomputing resources, as well as move data from data generation source such as scientific devices and IoT devices to computing resources and from computing resources to collaborators. Towards the realization D3 Center has been operating SQUID (Supercomputer for Quest to Unsolved Interdisciplinary Datascience) as a supercomputing system, ONION (Osaka university Next-generation Infrastructure for Open research and open innovatioN) as a data aggregation infrastructure. Also, RED (Research-EnhanceD) ONION as a high-speed data transfer service) has been explored. Furthermore, mdxⅡas a cloud infrastructure was installed at the University of Osaka and has been jointly operated by 9 universities and 2 research institutes.
In this talk the author will first explain the design, concept, and rationale behind each of these computing and data infrastructure components. Following this, the aturho will discuss how the convergence and operation of these infrastructures enhance the efficiency and productivity of scientific research Finally, the author will present a future vision for research infrastructure development, addressing challenges and opportunities.
From the background above, the D3 Center (former: Cybermedia Center) at the University of Osaka, which have a responsibility to provide supercomputing services to researchers in Japan, not limited to the University of Osaka, has been working on the development and deployment of a next-generation research infrastructure accelerating data-driven scientific research utilizing large-scale data. Through the realization of such a next-generation research infrastructure, we hope that it facilitates scientists and researchers to utilize our supercomputing resources, as well as move data from data generation source such as scientific devices and IoT devices to computing resources and from computing resources to collaborators. Towards the realization D3 Center has been operating SQUID (Supercomputer for Quest to Unsolved Interdisciplinary Datascience) as a supercomputing system, ONION (Osaka university Next-generation Infrastructure for Open research and open innovatioN) as a data aggregation infrastructure. Also, RED (Research-EnhanceD) ONION as a high-speed data transfer service) has been explored. Furthermore, mdxⅡas a cloud infrastructure was installed at the University of Osaka and has been jointly operated by 9 universities and 2 research institutes.
In this talk the author will first explain the design, concept, and rationale behind each of these computing and data infrastructure components. Following this, the aturho will discuss how the convergence and operation of these infrastructures enhance the efficiency and productivity of scientific research Finally, the author will present a future vision for research infrastructure development, addressing challenges and opportunities.