Research Computing as a Service

Provision of High Performance Computing clusters

We have multiple HPC clusters available with general services to facilitate high-intensity compute workloads if:

• You are running simulations, modelling, or large-scale data processing.
• You are training machine learning models.
• Your workloads exceed local workstation or lab capacity.
• Your projects require reproducibility and scalable computing.

RCaaS currently manages clusters available to all UofG staff and students:

Central services

• IT High Performance Compute clusters

College and school services

• MARS: MVLS Advanced Research System
• GES-Petrarch: School of Geographical and Earth Sciences

Access to this service

Proposal stage:

• Liaise with us to agree resource requirements (CPU and GPU hours, storage) and anticipated costs

Following notification of award:

• We can talk about the onboarding, training requirements and user access

Get in touch

Costs

These services are currently free, but future iterations of clusters will introduce charging based on resource usage / allocations.

• To account for these costs, the RCaaS team will facilitate a costing based on anticipated usage, currently expected to be in terms of CPU hours required, GPU hours required and storage needs.