This month, anything that doesn’t have me looking up to see if North Korea has lobbed a missile at the West Coast is a positive event. But this week, Intel responded to AMD’s Epyc launch with an epic launch of its own: the Purley version of its Xeon processor architecture. It clearly has come to play hard ball. Years ago, because things tended to be more generic, the processor played a far bigger role in servers and workstations. Today, a server can rely more heavily on the GPU than the CPU, can bottleneck on memory, storage, or internal transport rather than the processor more often, and just as often, must be purpose-built for whatever task it is being positioned in.
This is what keeps Open Power and ARM in the game because, with system tuning, systems based on these architectures are competitive for targeted workloads and there are a ton of different targeted workloads, allowing us a rather impressively wide selection of options. What makes this interesting is that the industry tends to fully lock down on a standard architecture, forcing all others out and while that was clearly the way it was going, we largely have reached equilibrium now. But I doubt this equilibrium will hold. War does that. AMD effectively declared war on Intel and Intel isn’t finding it at all funny this time.
But wars do interesting things to any segment, and this one is getting a lot of interest. Let’s talk about why this is a good thing.
Epyc vs. Xeon: The Wild Cards Are Off the Charthttps://o1.qnsr.com/log/p.gif?;n=203;c=204663295;s=11915;x=7936;f=201904081034270;u=j;z=TIMESTAMP;a=20410779;e=i
If you really start thinking about the market as it now exists vs. the one that existed when last AMD and Intel tangled for servers, a ton has changed. Back then, IT still largely controlled the data center, cloud services likely had more to do with shooting stuff into clouds to create rain than anything to do with computers, and GPUs were in gaming machines and workstations, for the most part.
Today, line of business largely dictates technology trends, the cloud is where most of the IT action really is, and across the board GPUs have invaded the servers of the world. The old war was relatively simple. This new one will be a complex mess, not the least of which will be the problem that the only folks who seem to be doing a great job of selling to line managers right now are those who have those pesky cloud services that have been disrupting this pace massively for the last few years.
Therefore, this time, both AMD and Intel seem uniquely focused on the cloud, but cloud providers will literally partner with anyone.
On top of this, we not only have NVIDIA as a non-aligned wild card in this mix, a firm that isn’t particularly fond of AMD or Intel, we have 3D Xpoint Memory, which Intel has branded Optane, but Micron also could sell to AMD. We have HPE, one of the biggest OEMs in the space, pushing an alternative architecture called “The Machine,” and we have IBM doubling down on Open Power (even mainframes are resurging).
We add to this the recent mega-merger between Dell and EMC to create Dell Technologies, arguably the most powerful technology company in terms of scale, and consider that the fastest growing market is China, where Lenovo has become the server player to watch. Things get interesting. assuming “interesting” means almost completely unpredictable.
But wars drive advancement and part of the industry’s growth problem is that too many of the firms were focused too much on cutting cost and not enough on innovation of late, which not only stalled the market but had a material adverse impact on the rate of productivity advancement.
This war will force these players to step up their game a lot, shifting from cost containment to investment, and that is the good thing I’m talking about.
Wrapping Up: Moving Toward Innovation Again
AMD’s and Intel’s battle is the big tip of an iceberg of coming conflicts as the IT industry continues to pivot toward line of business decision makers and efforts like AI, which will again redefine how companies do business. The moving parts are unprecedented and winners and losers will likely be more defined by who they are partnered with than what technology they bring to the table. King makers in this battle include Alphabet, Alibaba, Amazon, Dell Technologies, Lenovo and NVIDIA (which is doing surprisingly well now).
AMD’s path is easier than Intel’s because it simply needs to sell enough to be relevant to this space. But if Intel loses any share at all to AMD, it’ll take a significant hit. Intel’s interests have ranged far afield of late. I’d say the winner is the one that gets to the critical mass of partners, but given how fickle the cloud folks are, I’m pretty sure that’ll be a false indicator.
Regardless of how this one turns out, it’ll likely be one for the history books. From my perspective, I’m surprisingly okay with any war that doesn’t have me looking for missiles overhead. But it isn’t just a good thing because of a lack of missiles. It is a good thing because it is forcing those that provide technology off of cutting costs and back onto innovation, which the technology industry needs to live.
Rob Enderle is President and Principal Analyst of the Enderle Group, a forward-looking emerging technology advisory firm. With over 30 years’ experience in emerging technologies, he has provided regional and global companies with guidance in how to better target customer needs; create new business opportunities; anticipate technology changes; select vendors and products; and present their products in the best possible light. Rob covers the technology industry broadly. Before founding the Enderle Group, Rob was the Senior Research Fellow for Forrester Research and the Giga Information Group, and held senior positions at IBM and ROLM. Follow Rob on Twitter @enderle, on Facebook and on Google+