We can create high-performance computing clusters by utilizing Google Compute Engine VMs and Google Cloud Storage. The customer can run all the jobs using running HPC workloads of the application on Google's Cloud.
In the following diagram, you can see how we can have multiple Compute Engines (Compute Nodes) attached to a Compute Engine (HPC Head). At the same time, we are using multiple Compute Engines as a filesystem:
The compute portion of the cluster contains a Head Node running as orchestration software on a Google Compute Engine instance. The compute nodes and the related worker nodes are executing their task on Google Compute ...
