Many cloud providers are presenting units of measurement of resources that cannot be easily compared with their competitors, creating a pressing need for a single industry-wide standard.
Types of compute units not only vary between providers but also in what processing power they equate to. Although RAM size, measured in GB, is a common standard, the same cannot be said for CPU measurement.
For example, Amazon AWS uses Elastic Compute Units (ECU), which has an equivalent CPU capacity of a 1.0 – 1.2GHz 2007 Opteron or 2007 Xeon processor. However, Rackspace do not mention CPU allocations in their server configurations at all. My company uses vCPU, which has an equivalente CPU capacity of a 1.5Ghz 2010 Xeon processor. This lack of commonality is indicative of the cloud industry as a whole.
The problem has arisen because of the exponential increase in the demand for cloud services. Unfortunately, the industry has not been able to catch-up yet by bringing in common standards where customers need them most.
Until such standards become commonplace, cloud providers need to make sure that they are being transparent in communicating, in particular what they deliver for CPU capacity. Customers then need to benchmark providers against each other, ideally through independent third-parties, as well as making simple price comparisons.
Processor time is highly important for end users when dealing with CPU intensive applications. Developers in particular need to consider this when choosing a cloud infrastructure provider.
Developing common standards in data security, sovereignty and privacy is rightly occupying the focus of many in the cloud industry. However, we need complete transparency so that end-users can easily compare every part of a cloud providers’ market offer with their competitors.
Until we can all agree on a common standard, all cloud providers have a responsibility to be as clear as possible when communicating with their customers.