AMD on Tuesday announced availability of its processors on Amazon Elastic Compute Cloud (EC2). AMD EPYC-based systems will be used as web and application servers, backend servers for enterprise applications, as well as for test/development environments. The announcement is very important for AMD because Amazon Web Services is one of the world’s top adopters of servers.

AMD’s EPYC will be used for Amazon’s EC2 memory optimized and general purpose R5 and M5 instance families. AMD EPYC-powered M5 and R5 instances are offered in six sizes with up to 96 processors and up to 768 GB of memory. Meanwhile, AMD-based T3 instances will be available in 7 versions featuring up to 8 CPUs and 32 GB of memory.

AMD EPYC-based instances are already available in U.S. East (Ohio, North Virginia), U.S. West (Oregon), Europe (Ireland), and Asia Pacific regions. AWS intends to install AMD EPYC-powered machines to other datacenters in the future, so eventually EPYC-based instances will be available in other regions soon, Amazon said.

“The availability of multiple AMD EPYC processor-powered instances on Amazon EC2 instances marks a significant milestone in the growing adoption of our high-performance CPUs with cloud service providers,” said Forrest Norrod, senior vice president and general manager, Datacenter and Embedded Solutions Business Group, AMD. “The powerful combination of cores, memory bandwidth and I/O on AMD EPYC processors create a highly differentiated solution that can offer lower TCO for our customers and lower prices for the end-user. Working with AWS, the number one provider in cloud services, has been amazing for the AMD team and we are excited to see the new instances come online today for their customers.”

This is a breaking news. We are updating the news story with more details.

Related Reading:

Source: AMD

Comments Locked


View All Comments

  • Marlin1975 - Tuesday, November 6, 2018 - link

    Not really surprising. After the Meltdown and Spectre issues from Intel CPUs; Amazon and everyone else is hedging their systems so they are not fully reliant on just one supplier.

    If one performs better and/or offers better security/reliability they can use more of them. Easier to switch if you are familiar with more options.
  • BioHazardous - Tuesday, November 6, 2018 - link

    I actually recently was switched to a new server with my hosting company Linode, and I'm on EPYC processors on the new server. Though you don't typically get to choose what type of system you're on with them.
  • schujj07 - Tuesday, November 6, 2018 - link

    The sizing of the m5a & r5a instances doesn't make any sense. With 384 and 768GB of RAM respectively that means they are mixing DIMM sizes based on CPU sockets. Would have made more sense to have them be 512 & 1024GB, as well as 128 vCPUs but that is different issue.
  • Alexvrb - Tuesday, November 6, 2018 - link

    768 / 96 processors... looks fine to me
  • schujj07 - Thursday, November 8, 2018 - link

    1024/96 would have made more sense with the 8 channels of RAM per socket. 768 would be 8x64 + 8x32 instead of 1024 which is 16x64
  • zepi - Friday, November 9, 2018 - link

    What they allocate to you, doesn't necessary tell you anything about the underlying HW.

    Maybe they have a mixture of different hw and want to limit the maximum vm size to something that can fit comfortably to all different kind of HW configs.

    Or maybe they don't want to sell you 1:1 virtualization in these pricing bracket and rather have that as a separate product -> you can never buy whole host.

Log in

Don't have an account? Sign up now