AMD’s mission to add market share in the super-hot datacenter server processor market has taken its Epyc chip into the biggest datacenter of them all, Amazon Web Services, dominant leader in the public cloud services industry. Today at AMD’s Next Horizon conference in San Francisco, AMD and AWS announced the availability of the first AMD Epyc processor-based instances on Amazon Elastic Compute Cloud (EC2).
AMD said the new instances are available as variants of Amazon EC2’s memory optimized and general-purpose instance families. AMD-based R5 and M5 instances can be launched via the AWS Management Console or AWS Command Line Interface and are available today in the US East (Ohio, N. Virginia), US West (Oregon), Europe (Ireland) and Asia Pacific regions; availability elsewhere will follow soon.
AMD-based T3 instances will be available in the coming weeks. AMD-based M5 and R5 instances are available in six sizes with up to 96 vCPUs, up to 768 GB of memory. AMD-based T3 instances will be available in 7 sizes with up to 8 vCPUs and 32 GB of memory. The new instances can be purchased as on-demand, reserved, or spot instances on AWS, AMD 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.”
AMD, of course, has been mounting an aggressive price/performance competitive effort to grab business from Intel in the CPU server processor market. Although the latest numbers from Mercury Research put AMD’s market share at 1.3 percent, up from 0.5 percent earlier this year, the company has speculated that it expects to reach around 5 percent by the end of this year. Intel itself has said AMD could seize significant revenues in the sector; earlier this year in an interview with an Instinet industry analyst, former Intel CEO Brian Krzanich said the company’s objective is to limit AMD to between 15 and 20 percent of the server market.
Reached at the AMD conference, industry watcher Patrick Moorhead, president and principal analyst at Moor Insights & Strategy, said, “The AWS-AMD Epyc news will likely be the biggest news of the day. I don’t see this as replacing Intel at AWS, but rather AMD getting some business for the first time there. I expect Intel to keep doing the lion’s share of business there for the foreseeable future. This is huge news for AMD.”
For AMD, there’s no better cloud horse to jump on than AWS, whose public cloud services market share of above 35 percent is more than twice that of its nearest rival, Microsoft Azure. For the third quarter of this year, AWS cloud business revenues grew by 46 percent, and Amazon said it projects an annualized AWS run rate above $26 billion, compared with $18 billion for 2017. According to analyst group Synergy, quarterly cloud infrastructure service revenues were more than $17 billion.
“One thing our customers agree on is that they all like lower prices,” said Matt Garman, vice president, compute services, AWS. “Apart from adding to what is already the broadest and most capable set of compute services available in the cloud, these new AMD-based instances give customers an even lower priced way to run many of the most common applications.”
The AMD-AWS news adds to Epyc’s growing public cloud presence. Last month, AMD and Oracle announced availability on Oracle Cloud Infrastructure, following deployments on Azure, Baidu and Packet, a bare metal cloud for developers. In addition, SkySilk, a new public cloud provider that launched last month and provides Linux instances, also offers Epyc hardware.
In other AMD news today, the company announced the AMD Radeon Instinct MI60 and MI50 accelerators, which AMD said are the world’s first 7nm datacenter GPUs, designed for deep learning, HPC, cloud computing and rendering applications.
AMD said the accelerators provide ultra-fast floating-point performance and hyper-fast HBM2 (second-generation High-Bandwidth Memory) with up to 1 TB/s memory bandwidth speeds. They are also the first GPUs capable of supporting next-generation PCIe 4.02 interconnect, up to 2X faster than other x86 CPU-to-GPU interconnect technologies, and feature AMD Infinity Fabric Link GPU interconnect technology that enables GPU-to-GPU communications up to 6X faster than PCIe Gen 3 interconnect speeds, according to AMD.
A version of this article first published by HPCwire’s sister publication, EnterpriseTech,