Taiwan is set to launch its next-generation supercomputer in May, significantly advancing its artificial intelligence (AI) capabilities, according to Taipei Times. The system, equipped with Nvidia’s H100 GPUs, will initially provide 16 petaFLOPS of computing power, with plans to expand its capacity up to 200 petaFLOPS in the coming years.
This cutting-edge supercomputer is expected to accelerate the development of generative AI and strong AI, supporting breakthroughs in machine learning, natural language processing, and autonomous systems.
A Major Leap in AI Computing
The National Centre for High-Performance Computing (NCHC) is overseeing the project, ensuring that the supercomputer becomes a crucial resource for researchers across various fields, including physics, chemistry, mathematics, atmospheric science, engineering, life sciences, and AI. Researchers, academic institutions, and government agencies will be able to apply for access to the system.
Taiwan’s current AI supercomputer, Taiwania 2, built in 2018, has a computing power of nine petaFLOPS and has been widely used for AI tasks such as deep learning and data processing. However, with the growing demand for more advanced AI, the new system featuring Nvidia’s H100 GPUs will offer a substantial performance boost.
Expansion Plans & Future Goals
The NCHC aims to scale the supercomputer’s computing power to 100 petaFLOPS and eventually 200 petaFLOPS, depending on available funding. The system will support both generative AI—which produces text, images, and videos—and strong AI, which mimics human-like cognition, enabling advanced learning and reasoning.
Furthermore, the supercomputer will be integrated into an AI cloud service platform, featuring a traditional Chinese version of the Trustworthy AI Dialogue Engine. This platform will provide industry-wide AI services, allowing companies to develop customized AI applications and knowledge bases, reducing the cost of training large language models.
Taiwan’s Long-Term AI Ambitions
Beyond this new supercomputer, Taiwan’s NCHC also operates other high-performance computing systems, including Taiwania 3 and Forerunner 1. As part of Taiwan’s broader technological strategy, the government plans to enhance computing power and energy efficiency through initiatives like the Taiwan Chip-based Industrial Innovation Program and the Southern Silicon Valley project.
Taiwan aims to achieve a total of 480 petaFLOPS of computing power in the government sector by 2029, with national computing capacity expected to exceed 1.2 exaFLOPS, including contributions from the private sector, according to the Taipei Times.
This initiative positions Taiwan as a key player in the global AI and super computing industry, fostering innovation and technological leadership.