I had a chance recently to dig into the Google Cloud Platform, in particular Kubernetes clusters and virtual machine instances. This is the “Compute Engine” offering of the GCP, or Google Cloud Platform. The GCP also offers much more, for example, Cloud Storage [data, object storage], Cloud SQL [MySQL/PostgreSQL], and App Engine [building web + mobile apps].
I went into the Compute Engine section, fully understanding that Google has mega data centers filled with computers, hard disk drives and more virtualized resources spread throughout the world. These data centers are in Asia, Europe, and the USA, and because they are spread out like this, GCP is able to offer much in the way of reduced latency and fail-over capabilities. They mix in regions and zones within these areas to truly fight latency and provide fail-over, as seen below.