The ASUS Group has joined hands with Taiwan's National Health Research Institutes (NHRI) to advance supercomputer-powered medical research.
The official says the strategic collaboration will focus on three areas: computing power, data, and next-generation medical information. They aim to create a model similar to the UK's "Cambridge-1" in Taiwan.
AI-powered research
According to the NHRI, it first collaborated with ASUS in 2022 to establish "NHRI-1," applying AI technology to smart healthcare. In June 2024, the two parties signed a Memorandum of Understanding (MOU) on "Medical Information Technology and Biomedical Big Data Research and Application".
The joint initiative will focus on using medical big data, marking another big step in public-private partnerships in the AI age. Supercomputers combined with AI are already widely used for weather forecasting, drug development, and disease diagnosis.
However, some experts note that supercomputers face significant challenges in biochemistry, optimization problems, and cryptography, which may require multifunctional large-scale quantum computers. In other words, until quantum computing matures, supercomputers will continue to play a vital role.
Cambridge-1 v. Taipei-1
According to Nvidia, the UK's "Cambridge-1" is dedicated to accelerating healthcare research across medical imaging, genomics, and drug development.
Established by Nvidia in 2021, "Cambridge-1" was hailed as "the UK's most powerful supercomputer" at the time, offering over 400 petaflops of AI performance and 9.68 petaflops of Linpack performance. Its main purpose is to facilitate drug research, train models to understand chemical syntax and process digital pathology images.
Nvidia has established the "Taipei-1" supercomputer in Taiwan. 25% of its computing power will be allocated to academia and 75% to commercial use. Meanwhile, the National Science and Technology Council (NSTC) has commissioned the National Center for High-Performance Computing (NCHC) to build "Forerunner 1."
The supercomputer will have the optimal performance of approximately 3.5 petaflops. The machine, operational from June 24, 2024, is dedicated to research in physics, chemistry, mathematics, atmospheric science, engineering applications, and life sciences.
