CMGR means “compound monthly growth rate.” This CMGR metric shows the growth of each region of the CSPs we track. This measure of size is a factor of the number of instance type families offered in the region and the number of sizes offered in each type family (specific core counts and memory configurations). Because we scan these CSPs monthly, we are able to calculate changes in this metric from month-to-month.

Regions with larger CMRG are shown with larger icons. Positive and zero CMGR values are shown by a solid circle, negative CMGR values are shown by a solid square. Every active CSP region has a CMGR value.

We divide compute accelerators into four classes based on the architecture they use to accelerate math-intensive code. We show where these architectures are available for each CSP and region. See below for accelerator class definitions.

Circle size denotes the number of accelerator classes available in a region. A small empty circle indicates that no accelerators are present in that region.

We record many details about tracked CSPs’ IaaS accelerator options. We show accelerator chip vendor brands and general model numbers of the chips offered by tracked CSPs in each region. See below for a complete list of accelerator manufacturers and brands currently deployed by tracked CSPs.

Circle size denotes the number of accelerator brands available in a region. A small empty circle indicates that no accelerators are present in that region.

We record many details about tracked CSP’s IaaS processor options. We show processor chip vendor brands and the generations of those brands offered in each region. See below for a complete list of processor brands currently deployed by tracked CSPs and a list of which processor generation code names we compare to each other.

More processor brands available in a region are shown with larger circles.

Bare metal servers are specific instance sizes that reserve all of a physical server (the server “hardware”). Each instance type is based on a physical server specification–a specific model of processor that has a known number of cores with a specific total system memory that determines the type’s memory per core profile. Bare metal servers reserve the maximum specifications available for a type and are often offered without an operating system.

A large solid dot indicates that bare metal servers are available in that region. A small empty circle indicates that no bare metal servers at all are present in that region.

Storage-optimized instance types are specifically configured to provide compute services with storage capacity. GCP does not offer an explicit type family in this class, but storage-optimized type families are offered by AWS (H1, I3, D2, I3en types), Azure (L and Ls v2 types) and Alibaba Cloud (i1, i2, i2g types).

A large solid dot indicates that storage-optimized instance types are available in that region. A small empty circle indicates that none are present in that region.

Functions-as-a-Service (also referred to as “serverless”) is a server packaged as a logic execution engine, so the IaaS reservation is hidden by the PaaS service. All tracked CSPs offer functions as a service: Alibaba Function Compute, Amazon Lambda, Google Cloud Functions and Microsoft Functions / Microsoft Event Grid.

A large solid dot indicates that serverless functions are available in that region. A small empty circle indicates that no serverless functions are present in that region.

IoT Gateways are PaaS services designed to support bidirectional data transfers at scale, for streaming data from IoT sensors to a cloud and for smaller transactional management control and image update communications between IoT devices and a cloud. All tracked CSPs offer IoT Gateways: Alibaba IoT Platform, Amazon IoT Core, Google Cloud IoT Core and Microsoft IoT Hub.

A large solid dot indicates that IoT gateway services are available in that region. A small empty circle indicates that no IoT gateway services are present in that region.