RCC is currently deploying remote rendering on some of its HPC resources to allow researchers to visualise and interact with large, complex data sets and models with much higher performance than what is presently available.
RCC is working on infrastructure that will integrate Graphics Processing Unit-intensive applications into UQ’s cluster computing environments.
Remote rendering will allow data to be visualised ‘in situ’, meaning data does not need to be moved from the HPC onto local workstations for analysis.
“When dealing with very large data sets this will save time as there’s no need to transfer files and it will make it easier for researchers needing to interact with such data sets,” said RCC’s visualisation specialist, Oliver Cairncross. “Remote rendering can also provide advanced visualisation facilities to devices that aren’t equipped with the necessary hardware.”
GPU-equipped nodes are already available for numeric computation on the RCC/QRIScloud resource Euramoo (information about accessing these GPU nodes is contained in the Euramoo User Guide). Enabling remote rendering on these nodes will also provide accelerated visualisation capabilities. Once RCC’s work is complete, the GPU nodes can be used for both compute and rendering.
Early adopters are welcome to try RCC’s proposed remote rendering capabilities. Please contact Oliver Cairncross to register your interest: email@example.com.