The Hong Kong Science and Technology Park (HKSTP) and SenseTime have partnered to build a home-grown AI data centre [1].
This initiative seeks to reduce reliance on external computing resources by providing the local AI sector with high-performance capacity. By establishing a domestic infrastructure, the city aims to accelerate the industrialization of artificial intelligence, and support regional tech development [1].
The project will be developed in three stages [2]. The partners said the first phase of construction was expected to be completed by the end of 2023 [1]. This phased approach allows the facility to scale its capabilities as demand for AI processing grows within the region.
The long-term goal for the facility is significant. The partners said they have set a target computing power of 40,000 petaflops by 2030 [1]. This level of capacity is intended to support complex AI model training and large-scale data processing for various industries.
SenseTime, a Chinese artificial-intelligence firm headquartered and listed in Hong Kong, will provide the technical expertise for the build [1]. The data centre will be located within the Hong Kong Science and Technology Park, creating a hub for research and commercial application [2].
The collaboration represents a strategic move to bolster the city's position as a technology hub. By integrating the resources of a government-backed science park with the commercial capabilities of a leading AI firm, the project aims to bridge the gap between academic research and industrial application [1].
“The partners have set a target computing power of 40,000 petaflops by 2030.”
The establishment of a domestic AI data centre signals Hong Kong's intent to achieve greater technological sovereignty in the AI race. By targeting 40,000 petaflops of computing power, the city is positioning itself to attract AI startups and researchers who require massive compute resources without the latency or security concerns of using offshore cloud providers.
