AMD expects sales of its EPYC server processors to enterprise customers will soon reach the same growth rate of its overall server CPU business, which doubled in size last year
and more than doubled in the first quarter
with the launch of new processors.
That’s according to Dan McNamara, senior vice president and general manager of AMD
’s server business unit, who spoke about the chipmaker’s growth prospects for EPYC processors with hyperscale, high-performance computing and enterprise customers at Bank of America’s 2021 Global Technology Conference on Wednesday.
“We really see that this is starting to move,” McNamara said of EPYC sales to enterprise customers. “Today, we’re a bit underrepresented, but we believe that over the rest of this year and into 2022, that is changing. And we do believe that enterprise will grow at the regular clip that we’re seeing with the overall server business.”
That bodes well for AMD’s intensifying battle against Intel in the data center as EPYC sales grew so fast in the first quarter
that it marked the chipmaker’s highest single-quarter market share gain against Intel for server CPUs since 2006, when the company peaked in the market.
McNamara said he’s confident in AMD’s growth prospects with enterprises because the company is building trust with CIOs at major companies, including those in the Fortune 500, who are now talking to the chipmaker about their data center infrastructure goals for the next three to five years.
“It’s quite exciting to do that because when you win in the enterprise, there’s staying power, because you win their trust. It’s not just a transaction. It’s trust,” he said.
AMD’s growth in the enterprise server space is riding on the tailwinds of organizations modernizing their data centers, according to McNamara, for important workloads like virtual desktop infrastructure, which has benefited from an increasingly remote workforce. The value proposition of AMD’s EPYC processors have also closely aligned with enterprise needs for hybrid, multi-cloud environments and hyperconverged infrastructure, he added.
Another crucial factor for AMD’s growth in the enterprise is the company’s increased focus on solutions, or reference architectures, which makes it easier for customers to figure out the right configurations for a variety of workloads, according to McNamara.
“We’re bringing more and more solutions and more and more manpower to the table to enable and fine-tune and try and accelerate the bringing of AMD into their fleets on-prem,” he said.
AMD is making its biggest push into the enterprise this year with the launch of its third-generation EPYC processors
, code-named Milan, which are expected to go into more than 100 new server platforms from Dell Technologies, Hewlett Packard Enterprise and other OEMs.
That builds on the momentum AMD has made with hyperscalers like Amazon Web Services and Microsoft Azure, the largest buyers of processors in the market who were the company’s initial focus when it re-entered the data center with the launch of EPYC in 2017.
AMD expects EPYC to power more than 400 public cloud instances by the end of the year, but McNamara said the company is “much more represented” for internal workloads run by hyperscalers.
“When I talk to customers today, we have moved clearly from an alternative in the data center to a clear, strategic and trusted partner to most of the hyperscalers and the enterprise,” he said, adding that AMD has “become more of a clear part of their roadmap.”
The chipmaker’s footprint is growing fast within these “internal properties” because the high core counts and performance of EPYC’s processors allow cloud providers to maximize the number of virtual machines in a server rack to optimize total cost of ownership.
“We see a big growth within the internal properties as well as the public-infrastructure-as-a-service. But the internal properties, I believe, are a true measure of the value,” he said. “And the reason being is we have no control over how instances are sold by the cloud providers. But what we do have control on is the value we bring and the [performance] per dollar that we bring to the customer, and we do see these internal properties really jumping on the value prop of AMD going forward.”
AMD recently proclaimed that EPYC remains the “fastest server CPU on a per-core and per-socket basis” after the company was able to compare its latest Milan processors with Intel’s third-generation Xeon Scalable processors, code-named Ice Lake.
McNamara said the new 7-nanometer Milan chips will also compete with Intel’s next-generation Xeon Scalable processors, code-named Sapphire Rapids, which are expected to launch next year. But the company is also planning to launch its first 5nm EPYC processors, code-named Genoa, in 2022, which will give AMD even more ammunition against its top rival, he added.
“We’re going to have a nice one-two punch in the 2022 time frame with a leading 7nm and a leading 5nm product, so we feel very good about the stack coming in 2022 versus the competition,” he said.
An executive at a national solution provider said AMD has gained more interest and acceptance among his enterprise customers over the past year. While customers were initially interested in using EPYC processors for greenfield opportunities that involved smaller server deployments for emerging workloads, they are now more open to deals involving traditional data center infrastructure, he added.
“Today it’s much more an open dialogue than it was before, where they were like, ‘my whole environment’s Intel. Or my workloads are all set up this way. I can’t necessarily switch over to AMD,’” said the executive, who asked to not be identified because he was not authorized to speak publicly. “But now they’re looking at moving new workloads to the cloud and taking advantage of some of AMD’s influence there. It becomes a much easier conversation.”
There are multiple reasons it’s becoming easier for customers to consider AMD, the executive said. For one, the growing use of EPYC among hyperscalers lends a lot of validity to AMD in the enterprise space. It’s also a big deal that EPYC now has strong platform support from the OEMs.
“I think the ability to engage with the OEMs and diversify and broaden the platforms, the landscape becomes much greater,” he said.
AMD has also benefited from Intel’s stumbles with its 10nm and 7nm manufacturing processes, both of which suffered issues that resulted in products getting delayed, according to the executive.
“I think the climate continues to be challenging for Intel, and it allows for greater opportunities for conversation around AMD,” he said.
However, the executive said, it will take time for AMD to gain a much larger share of his company’s server deployments as Intel-based systems continue to dominate sales.
“It’s not going to go to 50 percent over the next year, year and a half, but many more opportunities are going to present themselves,” he said.
The executive credited AMD for hiring Hewlett Packard Enterprise veteran Terry Richardson as the company’s new channel chief and said the chipmaker is “leaning in and looking for unique partnering opportunities” in the channel, which will help the company’s enterprise efforts.
“I would say they’re being more prescriptive with their investments, and I think they’re tailoring their investments to the needs of whatever that [value-added reseller] or [national solution provider] is focused on,” he said. “They’re willing to step up and say, ‘We want to participate there.’”