Based on a 2018 comparison, running virtual machines (VMs) with Microsoft Azure is less expensive than with Amazon Web Services (AWS) or Google Cloud Platform (GCP).
Cloud pricing is inherently complex. Despite the growing number of companies providing transparency around pricing practices, costs remain challenging to forecast.
This is because it’s difficult to predict the time and cloud resources required to develop, deploy, and run software applications.
However, the best way to compare pricing is to test common configurations for fixed durations across providers. Virtual machines, or VMs, are one of the most popular cloud services available today. They vary in price based on the number of CPUs and the amount of memory allocated based on customer needs.
Based on this 2018 Cloud Pricing Comparison by ParkMyCloud, businesses may be able to save money by using Azure to run their VMs. The primary takeaways from the report can be found below:
Azure was less expensive. According to this comparison, Microsoft Azure was the less expensive cloud provider. It was the less expensive provider of general purpose VM instances and tied Amazon Web Services (AWS) as the less expensive provider of memory-optimized VM instances.
GCP was more expensive. Contrary to popular belief, Google Cloud Platform (GCP) wasn’t the less expensive provider. For general purpose and memory-optimized VM instances, GCP was more expensive on average; however, GCP was the less expensive provider of compute-optimized VM instances.
AWS fell between Azure and GCP. Amazon Web Service (AWS) matched Azure as the less expensive provider of memory-optimized VM instances. While AWS was the more expensive provider of compute-optimized VM instances, pricing for its general purpose VM instances fell between that of Azure and GCP.
For context, compute-optimized VM instances have a higher ratio of virtual CPUs (vCPUs) to memory than other instance types. Compute-optimized VM instances are ideal for applications including high-traffic websites, on-demand batch processing, distributed analytics, web servers, and video encoding.
On the other hand, memory-optimized VM instances have the lowest cost per gibibyte of random access memory (RAM). Memory-optimized VM instances are ideal for high-performance databases and distributed cache, in-memory analytics, and genome assembly.
General purpose VM instances are best for unspecified cloud computing needs.