Research IT Services
Microsoft Azure CycleCloud
PROS | CONS |
Easier Onboarding | Learning curve |
HPC system is already in place and configured | Users may need to invest time in learning how to use Azure CycleCloud and its various features. |
Support | Cost |
Supported by ESO Research IT for best practices in HPC | While Azure CycleCloud can help optimize costs, running HPC clusters in the cloud can still be expensive, especially for large-scale workloads. |
Flexibility | |
Azure CycleCloud supports multiple cluster types and schedulers, allowing users |
Researchers can use Azure CycleCloud to create, manage, and optimize high performance computing (HPC) clusters in the Azure Cloud. Azure CycleCloud provides various compute resources, such as CPU nodes and GPU nodes, for running research workloads with scalability and flexibility.
Pros | Cons |
Easier Onboarding | Availability |
Simple Web Based Allocation Request.
| This system is currently in the planning stages. Projected time of completion is estimated to be April 15, 2024. |
Support & Training |
|
Support and Training by ESO Research IT for best practices in HPC |
|
Hardware |
|
This system has new powerful HPC hardware that can deal with complicated calculations and big data fast. |
|
Cost |
|
Shared Condo Model for Nodes. A pay as you GO Storage system. |
|