Singapore’s first supercomputer for the healthcare sector is set to improve patient care and resource allocation.
Fully operational since July 31, the Prescience supercomputer lets National University Health System (NUHS) staff use AI to estimate the duration of a patient’s stay, allow better tailored treatments, and better allocate hospital resources.
At the click of a button, doctors may summarise patients’ case notes and write referral letters using locally trained large language models (LLMs).
Researchers can use medical big data to train AI models using multiple NVIDIA DGX A100 compute nodes that can accommodate the large sizes of LLMs, which was previously not possible with single graphical processing unit (GPU) systems.
As the world’s first five petaFLOPS AI system, NVIDIA DGX A100 is designed for all AI workloads, offering unprecedented compute density, performance, and flexibility. It features the NVIDIA A100 Tensor Core GPU to enable enterprises to consolidate training, inferenc, and analytics into a unified, easy-to-deploy AI infrastructure.
On December 3, 2021, NUHS signed an agreement with the National Supercomputing Centre (NSCC) Singapore to build this supercomputer dedicated to healthcare and medical research.
“Using the Prescience supercomputer, we are now able to train our own LLM for Singapore’s healthcare needs. From synthesising precise local medical knowledge to reducing the administrative work of our doctors and nurses, this LLM will bring benefits to both healthcare workers and patients,” said Associate Professor Ngiam Kee Yuan, Group Chief Technology Officer of NUHS.
“The Prescience supercomputer will significantly benefit Singapore’s healthcare research community and enable local healthcare professionals to develop tools that can increase the efficiency of healthcare delivery and accelerate healthcare innovations which ultimately benefit patients here,” said Bernard Tan, Director of Strategy, Planning and Engagement at NSCC.
Plans are under way to progressively roll out the LLM throughout the NUHS cluster.
Photo: Hush Naidoo Jade Photography
Entelechy Asia 