What is the UC Research Compute Hub?
The UC-RCH is a powerful platform where researchers can perform compute-intensive tasks. It鈥檚 accessed through a simple online user interface called (via HPCNow!). The platform is run by , a job scheduler that manages and allocates resources, optimising resource use and power consumption. UC-RCH is housed in our primary data centre on Dovedale, featuring 9 physical nodes and connecting to a virtual cluster with virtual machines and containers. The initial setup includes 1408 CPU cores and 16TB of memory, with the ability to expand to the AWS public cloud if needed.聽
What will this offer UC?
Flexible and reliable high-performance computing: UC-RCH is designed to support cutting-edge research with a focus on ease of use and access for our researchers.
Varied workload management: The platform manages highly variable workloads and can burst to the public cloud (AWS) for higher capacity if required.
Increased data security and robust disaster recovery: Ensures the safety and integrity of our research data.
Better collaboration: Allow researchers from different departments and institutions to work together more effectively.
The environment will be able to accommodate High Performance Computing (HPC) and High Throughput Computing (HTC) jobs via a standard workload scheduler (slurm). Examples of such works are simulation of black holes, fluid dynamics and weather simulations. We also have a few GPUs that can be used for Machine Learning or accelerating simulations when it is supported.聽
Researchers will also be able to access other high-level services through a web portal. Jupyter and RStudio environments which are a staple in many fields will be available. Those services will be expanded as we mature the platform. We already have a long list of services to enable, but we will take suggestions.
What's next?
The eResearch Team will be progressively rolling this out to more researchers across UC, while continuing to improve and release new functionality in response to researcher demand. Alongside this, we鈥檒l be completing the migration of existing Research Compute Cluster (RCC) users to the UC-RCH, which will enable the decommissioning of the old RCC.
Engagement and Support
The eResearch Team will collaborate with key stakeholders to establish effective solutions and patterns for different advanced computing use cases, that other researchers can then utilise and leverage.
We want to hear from you if you have a research funding bid that requires research compute or wish to find out more about the UC Research Compute Hub. This is an exciting opportunity for those keen to access and use powerful research computing!
A reminder also that regular eResearch Clinics are held every Tuesday at 11am in Puaka James Hight 210 鈥 please feel free to come and have a chat.
If you have any further questions, please don鈥檛 hesitate to get in touch at eresearch@canterbury.ac.nz.
