The data explosion in Science and Technology drives the demand for high-capacity
networks for accessing computers, storage, and collaboration; simultaneously, it brings
many restrictions on the generation, transmission, and utilization of research data. In
recent years, several institutes and universities have developed and utilized
ScienceDMZ-based big data super-highway to send and share scientific big data with
leading research institutes, universities, and institutions at ultra-high speeds.
Since 2018, APAN APRP WG utilizes NREN-based high-performance network infrastructure and
links the computing resources of each participating country to enable a high-performance
computing environment, thereby providing sufficient computing resources for members of
APAN and Asia regions. And also, these trends and efforts important data transmission
and sharing technologies in the era of big data, effectively share scarce computing
resources and promote breakthroughs in artificial intelligence technology. In addition,
this expands the scope of application not only to academic fields but also to various
fields.
To build a high-performance network environment based on APAN, ScienceDMZ technology is
used, and container cloud technology is used to build a high-performance computing
environment in a distributed environment. Especially, based on such a high-performance
network-based distributed high-performance computing environment would be support for
artificial intelligence, bioscience, earth science, and weather/climate which are not
only science areas but also social and life sciences fields. By connecting the idle
computing resources of Asian countries, we will form an APAN-based high-performance
computing environment, and serve as an infrastructure and platform to provide Asia
region and related projects and works in need of computing resources in the future.
To implement this, it is divided into three activities, and each will achieve the
detailed goals of (1) building a big data super-highway based on APAN (2) building a
distributed HPC platform, and (3) presenting pilot use cases. Activity (1) will
implement ScienceDMZ technology to improve the utilization performance of the physical
network based on APAN, while Activity (2) will implement a container cloud technology
system to the distributed HPC, relying on a high-performance physical network.
Activities (1) and (2) are applied to all participating institutions to establish big
data and high-performance computing transmission and sharing environments among
participating institutions. First of all, AI-based and bioinformatics-based science
research applications and use cases can be implemented, as part of Activity (3). These
applications are meant to improve the educational and research functions of the
participants.
Activity1. Building a big data super-highway based on APAN
Activity2. Building a distributed HPC platform as a research platform
Activity3. Presenting pilot use cases
|