Introduction to the Apollo GPU Cluster

Important

Access to the GPU Cluster is now available through the Researcher Dashboard (DashR). Please follow instructions on this page to submit your request.

The Apollo GPU Cluster is a high-performance research compute platform designed to accelerate AI workflows, scientific modelling, image processing, and other GPU-intensive research across the University of Sydney. This cluster is built with three NVIDIA DGX H200 nodes (24 H200 GPUs in total), high-speed InfiniBand networking, and 1 PB DDN EXAScaler parallel filesystem for fast shared storage.

This cluster is managed by the Sydney Informatics Hub (SIH). More information about the Apollo GPU Cluster can be found on the SIH’s Research Computing page.