If you think a car is four wheels and an engine, then a Porsche and a Hyundai are the same. When you are looking to purchase a car, you consider different variables and attributes like performance and price. And, while a Bugatti might have the best absolute performance, at $1M it’s arguably not the best price-performance.
What does this have to do with compute and storage cloud providers? Some think that compute and storage is now treated like a commodity, but much like shopping for a car, you have to consider different deliverables and attributes to make sure you’re getting what you want. You must also consider cloud suppliers. Is there anything that can differentiate them or are they all the same?
Let’s explore three cloud computing case studies and see if all cloud suppliers are really the same.
Use case #1: Finding a tie-breaker when two offerings cost the same
Let’s start with a big data application designed for European deployment. The application analyzes large amounts of insurance data and uses 654 virtual machines with over 195TB of storage.
The company wanted to model total price and price-performance over a 36-month period. Using the Burstorm application, which contains over 1,000 compute, storage, data center and network cloud service products in its real-time catalog, the best-priced cloud service was a tie between Amazon’s AWS and Google’s Cloud at approximately $7.5M. Microsoft’s Azure and Rackspace are noticeably more expensive.
While it was a dead heat between Amazon and Google on price, once you looked at price-performance, Amazon was better by about 30%. For this company, the price-performance attribute became their tie-breaker as they realized they would receive a better purchase deal.
Use case #2: Prioritizing price-performance ensures the best value
A company needs to do a large web deployment with 330 virtual machines and 15.7TB of storage to be used to deliver the results of scientific modeling data, which was to be deployed in North America. Furthermore, this application is a mix of Windows and Linux. The enterprise cloud architect modeled a 12-month term.
For this organization, Google was the price-performance leader followed by Rackspace and Amazon. Based on the data, IBM’s cloud was very competitive on absolute price, but in price-performance fell short of the other three offerings. Large projects can be costly but by broadening their cloud supplier criteria beyond price alone, this company was able to secure the best deal for its project.
Use case #3: Cut the complexity with effective option analysis
The third case is a small web application - 18 virtual machines and a little over 1TB of storage to be used for consumer market place. Designed to be deployed in the western United States, the cloud architect modeled a month-to-month use case. These circumstances opened up a whole new set of options.
With more choice comes more complexity, so before making a commitment, the company performed an analysis of the available cloud suppliers in order to determine the best option. Linode, a Linux-only player, ended up on top, having both best price and best price-performance versus other cloud service providers. Google and AWS were in a virtual tie for second place.
So just like all automobiles are not the same, all cloud suppliers are not the same. Cloud services are becoming more and more specialized which can help buyers differentiate offerings and prioritize based on their specific needs and projects. To learn more, watch the Compute & Storage Cloud Services module.