Singapore’s National University Health System (NUHS) is working to use edge supercomputing in order to speed up the training of artificial intelligence (AI) models in healthcare.

Indeed, researchers at NUHS are collaborating with Singapore’s National Supercomputing Centre (NSCC) to build this petabyte-scale edge supercomputing infrastructure named Prescience. It will then be used to predict patient health trajectories, healthcare robotics, and genomics variant calling.

Moreover, the supercomputer will be able to train conversational chatbots so that they can converse with people in a more natural way and advise patients on diseases and care. In time, the supercomputer will also allow NUHS to start AI projects at scale by anonymising data used to train AI models.

In order to achieve that, Prescience will be powered by graphics processing units, which is faster than anything and will help reduce training times to hours and ultimately, improve patient care.

Prescience is planned to be ready by mid-2022 and will be residing at NUHS’s premises so as to address data privacy and security requirements.