AWS News Blog

New Lower-Cost, AMD-Powered M5a and R5a EC2 Instances

Voiced by Polly

From the start, AWS has focused on choice and economy. Driven by a never-ending torrent of customer requests that power our well-known Virtuous Cycle, I think we have delivered on both over the years:

Choice – AWS gives you choices in a wide range of dimensions including locations (18 operational geographic regions, 4 more in the works, and 1 local region), compute models (instances, containers, and serverless), EC2 instance types, relational and NoSQL database choices, development languages, and pricing/purchase models.

Economy – We have reduced prices 67 times so far, and work non-stop to drive down costs and to make AWS an increasingly better value over time. We study usage patterns, identify areas for innovation and improvement, and deploy updates across the entire AWS Cloud on a very regular and frequent basis.

Today I would like to tell you about our latest development, one that provides you with a choice of EC2 instances that are more economical than ever!

Powered by AMD
The newest EC2 instances are powered by custom AMD EPYC processors running at 2.5 GHz and are priced 10% lower than comparable instances. They are designed to be used for workloads that don’t use all of compute power available to them, and provide you with a new opportunity to optimize your instance mix based on cost and performance.

Here’s what we are launching:

General Purpose – M5a instances are designed for general purpose workloads: web servers, app servers, dev/test environments, and gaming. The M5a instances are available in 6 sizes.

Memory Optimized – R5a instances are designed for memory-intensive workloads: data mining, in-memory analytics, caching, and so forth. The R5a instances are available in 6 sizes, with lower per-GiB memory pricing in comparison to the R5 instances.

The new instances are built on the AWS Nitro System. They can make use of existing HVM AMIs (as is the case with all other recent EC2 instance types, the AMI must include the ENA and NVMe drivers), and can be used in Cluster Placement Groups.

These new instances should be a great fit for customers who are looking to further cost-optimize their Amazon EC2 compute environment. As always, we recommend that you measure performance and cost on your own workloads when choosing your instance types.

General Purpose Instances
Here are the specs for the M5a instances:

Instance Name vCPUs RAM EBS-Optimized Bandwidth Network Bandwidth
m5a.large
2 8 GiB Up to 2.120 Gbps Up to 10 Gbps
m5a.xlarge
4 16 GiB Up to 2.120 Gbps Up to 10 Gbps
m5a.2xlarge
8 32 GiB Up to 2.120 Gbps Up to 10 Gbps
m5a.4xlarge
16 64 GiB 2.120 Gbps Up to 10 Gbps
m5a.12xlarge
48 192 GiB 5 Gbps 10 Gbps
m5a.24xlarge
96 384 GiB 10 Gbps 20 Gbps

Memory Optimized Instances
Here are the specs for the R5a instances:

Instance Name vCPUs RAM EBS-Optimized Bandwidth Network Bandwidth
r5a.large
2 16 GiB Up to 2.120 Gbps Up to 10 Gbps
r5a.xlarge
4 32 GiB Up to 2.120 Gbps Up to 10 Gbps
r5a.2xlarge
8 64 GiB Up to 2.120 Gbps Up to 10 Gbps
r5a.4xlarge
16 128 GiB 2.120 Gbps Up to 10 Gbps
r5a.12xlarge
48 384 GiB 5 Gbps 10 Gbps
r5a.24xlarge
96 768 GiB 10 Gbps 20 Gbps

Available Now
These instances are available now and you can start using them today in the US East (N. Virginia), US East (Ohio), US West (Oregon), Europe (Ireland), and Asia Pacific (Singapore) Regions in On-Demand, Spot, and Reserved Instance form. Pricing, as I noted earlier, is 10% lower than the equivalent existing instances. To learn more, visit our new AMD Instances page.

Jeff;

PS – We are also working on T3a instances; stay tuned for more info!

 

Jeff Barr

Jeff Barr

Jeff Barr is Chief Evangelist for AWS. He started this blog in 2004 and has been writing posts just about non-stop ever since.