Research Computing in the Cloud

Research computing is not just about HPC. Researchers need access to storage and virtual machines pre-installed with their scientific applications. OCF provides solutions that allow our customers to create a centrally managed public cloud environment just for this purpose

Research Computing in the Cloud

Providing a research computing infrastructure that meets the needs of all your users has become an increasingly complex task. OCF can build temporary research computing environments in the cloud. Whether you are looking to test some of the latest hardware or to build temporary cloud based workstations for a conference or training course, OCF can create an environment to meet your needs. 

Research Computing - Cloud Use Cases

The public cloud brings a huge amount of flexibility to our research customers, giving you access to the latest technology, at practically any scale, as and only when you need it.

Public cloud providers offer:

  • The latest hardware including GPU’s and CPUs
  • High Performance Interconnects
  • Pre-configured hosted solutions are available with OS’s and even specific application stacks including Machine Learning, Data Management, IoT etc pre-installed.
  • Ability to right-size your environment for your requirement today

 

OCF have been involved in a number of research computing cloud projects including:

  • Configuring an environment in GCP that gave each attendee access to a GPU workstation specification server, pre-configured with an imaging software stack for the length of a 3 day conference and ‘modelathon’. This solution included shared storage, a licensing server and on-site support for the length of the conference.
  • Building a research computing cloud portal to give researchers an approved method of requesting Microsoft AZURE cloud resources using Red Hat Cloudforms. This solution gave the customer a limited catalogue of service offerings, an estimated maximum cost prior to requesting a service and an approval stream that enables budget holders to approve or deny a request for cloud resources.
  • Demonstrating HPC Cloudburst using Slurm and AWS.

Request a call-back to discuss if your Research Computing workload is suitable for the public cloud