Gwangju City Launches Pilot Service for 'National AI Data Center'

Free Support for AI R&D Resources to 11 Companies Until the 31st

[Asia Economy Honam Reporting Headquarters Reporter Park Seon-gang] Gwangju Metropolitan City and the Artificial Intelligence Industry Convergence Project Group announced on the 2nd that they have started the pilot service of the "National AI Data Center," which is being promoted as part of the "AI-Centered Industrial Convergence Complex Development" project.

The National AI Data Center will be established in the Advanced District 3 with a total project cost of about 90 billion KRW, boasting world-class performance with computing power of 88.5 PF (petaflops) and storage capacity of 107 PB (petabytes).

Here, companies engaged in AI-related learning, inference, and analysis research, product development, and service launching, including AI SMEs, public institutions, and universities, will be supported. Priority support will be given to AI companies located in Gwangju and AI companies planning to relocate their headquarters, branches, or research institutes to Gwangju.

Additionally, an AI accelerator testing and application environment will be established and provided, with plans to offer an environment for AI semiconductor verification in the future.

In particular, to immediately support companies that urgently need expensive AI accelerators, from April, NHN-operated cloud will first provide 8.85 PF accelerator performance and 10.7 PB storage space.

To this end, Gwangju City and the project group will select about 100 companies and support up to 125 TF and 50 TB of data storage per company. From 2023, through the National AI Data Center established in Gwangju, they will provide 88.5 PF performance and 107 PB of storage space to about 1,000 companies.

This pilot service was promoted to check AI development environments such as AI accelerators, data lakes, and software solution services (SaaS) ahead of the official service of the "National AI Data Center," which will start on April 1.

Gwangju City and the project group selected 11 institutions based on the urgency of task execution and research plans among those who submitted utilization applications and plans through a preliminary survey in December last year, and will provide up to 125 TF (8 GPU) performance and 10 TB storage space per company until the 31st.

During the pilot service period, they will test the utilization environment and functions provided by the future National AI Data Center, including computing resources, software solution services, and data lake services, through usage patterns and utilization checklists, including related education and support services, and evaluate user satisfaction.

Any shortcomings will be improved in consultation with NHN and reflected in the official service in April.

The official service starting in April will prioritize support for companies (institutions) related to the AI development project and selectively provide services to other AI research and development companies.

Companies can utilize computing resources and development environments provided by the project group and data secured and shared with users to enable full-cycle linkage from AI research and product development to service launching.

Also, through the flexible configuration features of the cloud environment and the 20 PF-class HPC-AI environment supported from 2023, learning and analysis time will be drastically reduced to maximize business performance.

The support period is up to four years, expected to be a great help to companies aiming to secure product competitiveness through stable AI technology development.

Institutions and companies wishing to use the official service can apply by filling out the utilization application and plan forms announced on the project group’s website at the end of March. For more details, contact the Artificial Intelligence Industry Convergence Project Group.

This content was produced with the assistance of AI translation services.