EC2 moves to vCPU approach. Does it really affect us?

AWS recently moved from ECU approach to more traditional vCPU approach for all their EC2 instances. Firstly I would like to discuss about the concept of ECU and vCPU and secondly I would throw some light on how this change would affect existing customers and the new customers.

ECU & vCPU

In order to make it easy for developers to compare CPU capacity between different instance types, Amazon had earlier defined EC2 Compute Unit (ECU). One EC2 compute unit provides the equivalent CPU capacity of a 1.0-1.2 GHz 2007 Opteron or 2007 Xeon processor. A virtual CPU (vCPU) also known as a virtual processor, is a physical central processing unit that is assigned to a virtual machine.

EC2 is shifting from old generation instances i.e.  general instances (m1), compute optimized instances (c1), memory optimized instances (m2) and storage optimised instances (HI1) to new generation instances namely general instances (m3), compute optimized instances (c3), memory optimised instances (r3) and storage optimised instances (i2, hs1). Most new generation instances have SSD drive as instance storage except micro instances. New generation instances have latest Intel Xeon E5-2670v2 processor except hs1.8xlarge instance which has Intel Xeon E5-2650 processor. The clock speed of the vCPUs for the new generation instances, can be viewed while launching an instance on EC2 as shown in the below image.

Advantage or Disadvantage

For the existing customers who are using AWS from a long time and at a large scale, there could be a need for change. They have been working with ECU and might have standardized the concept of ECU internally over the years. Now existing customers might have to adapt to the more traditional vCPU approach and have to manage their instances accordingly. This would definitely mean a change in the way the organization has been dealing with EC2 hardware planning and provisioning, but not a difficult change.

But for the new customers, I think it will be very useful because vCPU approach is more traditional and is used by many cloud providers. As all major cloud providers like VMware, Bluelock, Rackspace, CenturyLink are using vCPU for their machines. Hence, it will be easier for new customers who want to migrate to AWS EC2 to understand current EC2 hardware specs.

Overall, I think unless some existing customers really want to understand vCPU approach, this change would not affect them. But moving to vCPU approach will surely bring in new customers to AWS as it is now easier to decide among different instances (machines) available and even the customers who are using other cloud service can easily migrate to AWS. So, what’s your opinion? Please do leave a comment below.

WRITTEN BY CloudThat