AMD Rome EPYC processors represent the foundation of AMD's strategic shift to compete in enterprise data center markets.
AMD's diverse product portfolio functions like a multi-cylinder engine, with different business lines operating at varying intensities. While the custom processor business for Sony and Microsoft game consoles proved somewhat soft during the second quarter, all focus is on the second-generation "Rome" Epyc processors for servers. Chief executive officer Lisa Su revealed during a conference call that the Rome announcement is set for August 7, describing this as what matters more than anything AMD has done in the past decade and a half. Unlike the Great Recession a decade ago when AMD ceded the datacenter CPU business to Intel and left the datacenter GPU market open for Nvidia, AMD under Su has gotten its CPU and GPU acts together and is now generating approximately 15 percent of its revenues from products aimed specifically at the datacenter.
In the second quarter, AMD's datacenter business was weighted more heavily toward CPUs as the company pushed first-generation "Naples" Epyc processors into the Amazon and Microsoft public clouds and conducted initial customer shipments of Rome processors to unnamed hyperscalers and cloud builders. Even with datacenter GPU sales rising, server CPU sales rose faster, and the CPU share of the AMD datacenter business grew. Su indicated that in the second half of 2019, CPUs will get an even larger slice of the pie as Rome ramps further and ships in volume. However, Su worked to temper expectations by noting that AMD has observed the same tepid spending among hyperscalers and cloud builders that has affected Intel in recent quarters, despite Intel shipping new "Cascade Lake" Xeon SP processors launched in April. "Our plan was always very heavily second-half weighted," Su explained on the call. "And from our standpoint, we have seen significant customer engagement and pull for the Rome product and we see a number of new installations that will come online over the next couple of quarters. So I believe that there is some cloud digestion that is happening out there. I also believe that given where we are from the product cycle standpoint, we are well positioned to grow."
AMD has secured significant wins with its Epyc processors and Radeon Instinct GPU accelerators, notably the custom CPUs and GPUs going into the 1.5 exaflops "Frontier" supercomputer at Oak Ridge National Laboratory—displacing both IBM's Power10 processor and Nvidia's future Tesla GPU accelerator—with 2021 deployment. The Rome chip features eight chiplets, each with eight Zen 2 cores etched in 7 nanometer processes from Taiwan Semiconductor Manufacturing Corp, wrapped around a central I/O and memory controller chip etched in the established 14 nanometer processes from GlobalFoundries, creating a package with a maximum of 64 cores, eight memory controllers, and 128 lanes of PCI-Express 4.0 peripheral bandwidth. Su stated: "Rome is on time and exceeding expectations, delivering leadership performance and TCO across a significantly expanded number of cloud and enterprise workloads. Customer excitement continues to grow. Compared to our first-generation Epyc processors, we have more than twice the number of platforms in development with a larger set of partners. We also have four times more enterprise and cloud customers actively engaged on deployments prior to launch. As a result, Rome is on track to ramp significantly faster than our first-generation Epyc processor. We are seeing particular strength in HPC where we offer leadership compute density and I/O."
Rome delivers twice the integer performance and four times the floating point performance, clock for clock, compared to Naples, and is driving twice as many platforms and four times as many active customer engagements—a numerical synchronicity that is not causal. On the cusp of the Rome launch, AMD's datacenter products are probably driving around $230 million. Analysis of the business indicates that if Naples has been driving 60 percent of the formal datacenter business in 2019 so far, this represents approximately $114 million in Epyc processor sales in Q1 (up by a factor of 6X year-on-year) and approximately $138 million in Epyc CPU sales in Q2 (up by 3.6X compared to the year-ago period). These results, achieved against increasingly tough comparisons as AMD captured server market share in 2018, suggest it is absolutely conceivable that AMD could soon be driving a $1 billion annualized run rate for server processors alone and hitting a $2 billion annual run rate.