Investors who own Advanced Micro Devices (NASDAQ:AMD) though, or are at least thinking about adding a position in AMD stock, would be wise to add the word ‘hyperscale’ to their lexicon. Hyperscale is the practice of efficiently going from a few servers up to thousands.
While not a new premise, it’s about to become a monumentally important one to the cloud computing world. Advanced Micro Devices is positioned to not only usher in that era, but help shape it.
It will have to, in fact, if it wants to take more market share in the burgeoning market.
AMD Stock and Hyperscaling
Some tech-savvy people know of its enterprise-level wares, but by and large the majority of investors know and love AMD through their consumer-minded lens.
That is, the company makes graphics cards and computer processors that video gamers love, and the cryptocurrency boom was wildly beneficial for AMD stock, even if it was a short-lived boon.
There’s a whole side of Advanced Micro Devices, however, that most investors at least somewhat look past, unsure of what exactly it is.
Big mistake. It’s this other side that will actually drive AMD stock higher or lower, in the long run.
That other side of the company’s business is data center solutions. Last quarter, this segment saw more than 8% revenue growth, reaching a total of $939 million, or roughly two-thirds of total sales. Though not all of this arm is necessarily data center-driven, a huge chunk of it is.
The next evolution in the arena? The normalization of so-called hyperscale data centers. Advanced Micro Devices is moving into a good position.
There’s no hard-and-fast definition for ‘hyperscale,’ though most industry experts would describe it as the ability to quickly and easily scale-up the capacity of a data center, using low-cost components that make that scale-up economical.
Where virtualization of hardware is possible, that’s often the preferred course. Data center operators love its low cost and flexibility, which have become critical within the fast-moving and ever-changing arena.
It’s a sweet spot for Advanced Micro Devices’ Epyc chip, a processor built from the ground up to effectively work in a hyperscale cloud computing environment even before the idea has fully gelled.
Epyc Ideal for Hyperscaling
The company’s first Epyc microprocessor, a CPU designed with server networks and data systems in mind was released in mid-2017.
The rival Xeon series from Intel (NASDAQ:INTC) was getting a bit long in the tooth at the time, and the marketplace was ready for an upgrade. The early editions of the Epyc CPU were reasonably well received.
Its early success, however, was only a prelude to what was in store.
That success would be tough to measure given the data so far. Near the end of last year is was estimated that AMD only controlled 2% of the server chip market. The rest of it remained hogged by Intel.
That 2%, however, is up from 0% a year and a half earlier, and stolen from the king of data center hardware in a market where customers are slow to make purchasing decisions, and slow to switch providers.
The company, to the delight of AMD stock owners, is talking about capturing 5% of that market in the foreseeable future.
The upcoming launch of the ‘Rome’ version of its Epyc chip (tech companies assign code names to everything, though they’re rarely a secret) may well get Advanced Micro Devices over that hump.
It will fit and work in the same boards that used earlier versions of the processor, but will introduce the oft-discussed 7 nanometer leap and up to 64 cores to the mix.
In plain English, it will arguably be the most powerful data center chip on the market, and at a very low-cost in terms of cost-per-capacity.
It’s also the chip that could finally convince technology companies hyperscaling is worth the switching cost and effort. As Patrick Moorhead, principal at Moor Insights & Strategy, recently explained of Rome:
“What makes it so attractive is that it’s not just AMD selling something for less. It’s that a single socket server with all of the bandwidth and cores that are available will allow people to make smaller servers so you can have a higher density, and density is key particularly with the hyperscalers or even people in hosting.”
The soon-to-launch updated Epyc chip gives AMD an even more powerful weapon to wield in what will soon be an $80 billion market, up from $25 billion in 2017, now that corporate customers can see a clear cost-effective upside.
Looking Ahead for AMD Stock
There’s still work to be done. Intel isn’t sitting on its hands, and Nvidia (NASDAQ:NVDA) still dominates the GPU market where Advanced Micro Devices competes. AMD is also still conspicuously missing from the artificial intelligence arena, where Nvidia rules and Intel at least keeps Nvidia honest.
Nevertheless, penetration of the data center market and the hyperscaling sliver of that market in particular is not only very possible for Advanced Micro Devices, it’s very likely. GPUs, as well as AMD has carved out a small but loyal following, aren’t going to be a key growth driver.
Bottom line? Investors in AMD stock should take note of Epyc’s growing role in the young hyperscaling market, even though it’s unclear what that landscape will actually look like a year from now.
It’s a development that’s still not been given due attention.
As of this writing, James Brumley did not hold a position in any of the aforementioned securities. You can learn more about James at his site, jamesbrumley.com, or follow him on Twitter, at @jbrumley.
