It looks like the end of an era is upon us, today Intel announces the phase-out of their branded motherboard line. While this may be seen as a negative, SemiAccurate sees this re-prioritization of resources as a good thing for them and the market.
Short story, over the next three years, Intel is going to wind up their desktop mobo business, and refocus the design resources on making FFRDs. FFRD stands for Form Factor Reference Design, what amounts to the concept cars of the PC industry. Some times you end up with a Buick Y-Job, sometimes you end up with an Edsel. Or a NUC, but that looks nothing like the first Mac-mini. Then again, quad screens went nowhere fast, but the Metro became the MacBook Air so there is quite a bit of merit in making FFRDs.
So that is what Intel is doing. Over the next three years, the Intel branded desktop motherboard line is going away. Nothing is likely to be cancelled, given the design horizons of these things, it is more likely a case where anything that has been started now will be completed, but nothing new will be green lighted. There won’t be a major shift in anything, it will be gradual and painless. If Intel puts out new boards for the next two years, and sells the last few for a year, that would be all three years, so Sky Lake is likely to be CPUs only as far as Intel is concerned.
Update 1/24/13 @11am: Clarified exit to indicate only desktop, not server, boards are winding down.
As the teams finish up one project, they will likely be placed on another, but instead of the glamourous life that is building a mid-range corporate Q-series desktop board, they get to design a flashy new thingy that wows people in keynotes. What that thingy is, and what it needs, is the design and engineering process that leads to an FFRD. That in turn may lead to a Mac Mini, MacBook Air, Monorail, or in the case of the NUC, hopefully nothing. Come to think of it, the FFRD path is the glamourous life, at least in comparison to a down-costed vPro mini-ITX board.
So overall, it may seem like a big deal and a big withdrawal, but in reality it isn’t. It is going to be a slow, planned transition that will allow Intel to focus more on what it actually needs to do, make markets for it’s products. Given how few prospects they currently have, SemiAccurate only wonders why these steps weren’t take long ago. This is a good thing for Intel.
The real interesting question is what this does to the market. Given the size of Intel’s branded board sales this has some impact on the other motherboard makers. Intel also has a vastly disproportionate mindshare in several key markets, and now those are about to be thrown up for grabs. Lets take a look at how this all shakes out.
The motherboard business in recent years has shrunk to only two major players, lets call them Tier 1 vendors, Asus and Gigabyte. Both of these have sales of about 20 million units per year. Moving down a step is Asrock with about 7-8 million sales units a year, not even half of the big two. You could argue that ECS and MSI, both of which have sales significantly less than Asrock are tier two or three, but it doesn’t really matter. When the Asus purchase of Asrock goes through, insiders say it is still on track, there will just be Asus, Gigabyte, and other.
However you slice the market up, Intel is not a significant player in terms of volume. Their unit sales are about four million a year, roughly tied with ECS for last on the top five chart. To put that in perspective, it is a low single digit percentage of the desktop market. This does vary greatly by geography, in some markets, specifically a few South American countries, Intel commands overwhelming marketshare.
No matter what they sell for volume, Intel does have an overwhelming mindshare among the average consumer. Intel makes the chips, so they must make the best motherboard, right? That isn’t always the case, and enthusiasts don’t flock to Intel branded boards, but in some markets like business and government sales, they are the 800 pound gorilla. In spite of their advantages in mind-share and cross-selling, (legal) kickbacks and perks are huge in the channel, Intel board sales never took off.
Then again, four million units isn’t exactly peanuts, and it is likely very profitable. Intel has to provide reference designs to many players no matter what, so selling those to consumers is a fairly low bar to hurdle. Most people believe that their entry in to the branded board market in the first place was based more on someone saying, “Why not?” in a bar after work than a careful plan for world domination.
When Intel leaves this market, it will be a blip, not a yawning void left after an earthquake. Other players are currently many times their sales, and are holding their own against Intel’s advances. If it does change anything substantial, leaving this market will remove a contentious sore point with some of their largest customers. Intel is no longer competing with the mobo vendors, so the nagging fears that underlie any strategic talks and plans are aiding the enemy are no longer a concern. Or will no longer be in three years. For the little fish, it also removes one more competitor, not that this will matter much at their end of the market.
One area that will see a disproportionate shift is the large business and government sales area. These customers are big on stability, security, and similar metrics. Here Intel is seen as the ‘safe bet’ in the same way that no one ever got fired for buying IBM. This is far more perception than an actual quality concern, but it is real. These are very lucrative markets that were dominated by Intel, and now they are wide open. The big two can go after this business harder than ‘other’, so if anything, their margins should benefit a little.
Looking at the big picture, Intel phasing out motherboards sounds like a huge deal. Looking at the numbers, other than for a few minor markets, it simply is not. Looking farther out at the timelines, a three year sunset means nothing drastic is going to happen, and teams designing boards will be moved to FFRDs, a task that uses the same skill sets, but is much more interesting work. Partners will lose a competitor, and close ties with Intel on strategic topics will be viewed with far less suspicion than before. For Intel, this is a win/win scenario, there is no real down side, and a lot of potential gains to be had. No complaints from this peanut gallery.S|A
Latest posts by Charlie Demerjian (see all)
- Thing go bump(gate) in the night for Nvidia’s GP100 Pascal GPU - May 3, 2016
- Cavium’s Octeon TX blends compute and packet moving - May 2, 2016
- Qualcomm releases Zeroth API to developers - May 2, 2016
- Another detail about Qualcomm server SoCs revealed - Apr 27, 2016
- AMD finally really honestly launches the dual Fury - Apr 26, 2016