Should Intel Make ARM Processors?

Right now, Intel is losing out on the shift to mobile computing. Sure, there are devices that use Intel’s mobile processors, but it seems that ARM is the leader. The heavyweights of smartphones, Apple and Samsung, both use ARM processors, and those two combined represent nearly 60% of all new smartphones sold. So why wouldn’t Intel include an ARM processor in their lineup? I think it makes sense, and it allows Intel to leverage their advantages while having a way into the growing mobile market. I’ve written about this before, but in a slightly different context. I thought it was time to explore this in more detail, and specific to Intel.

No Imminent Danger, But…

First, let’s be clear  on something. While mobile computing is on the rise, the traditional PC is still chugging along. Most of the 400 million non-tablet PC’s sold in 2012 will have an Intel processor in them. A few will have an AMD processor. But Intel still controls the lion’s share of the non-tablet PC market (which includes desktop, laptops and servers). It isn’t as if Intel is in imminent danger, and it is entirely possible that Intel could see a boost in processor sales for mobile devices. Intel, though, is on the outside looking in when it comes to being part of the smartphone and tablet market right now, and tablets are cutting into PC sales (despite what some may want to claim). If the trend continues, this will start to show up significantly on Intel’s bottom line. Not tomorrow, or even next year, but in the next few years.

It is unlikely that Apple or Samsung will use Intel’s x86-based Atom processors for their smartphones. Why? Because Samsung already has in-house manufacturing capability for ARM, and Apple is designing it’s own ARM chips. Both have heavily invested in the platform, and are unlikely to abandon it. That leaves Intel trying to convince the remaining smartphone manufacturers (who account for 40% of the market and shrinking) and tablet manufacturers that aren’t iPad (again, 35-40%, but a lot of those aren’t using Intel now) to use their wares.

Intel Has The Resources and Skills

There is no reason why Intel couldn’t design and build an ARM processor, and a good one at that. The company certainly as the engineering talent. They have the facilities to make them economically and reliably in large quantities. When it comes to designing and building ARM, on paper Intel has what it takes. All they would need is a licence, and they may already have one of those.

Intel is, ultimately, an IC and microprocessor company. While they are most strongly identified with the x86 chip, it isn’t the only thing in their inventory, and it isn’t the only thing they put their time and effort into. There was a commercial once about a young executive talking about a company investing in ethanol. One of the older executives chimes in with “do I need to remind you we are an oil company?” as they tried to shoot down the idea. The executive’s response was classic: “no, we are an energy company”. Intel isn’t an x86 CPU company. They are an integrated chip and microprocessor company. Making ARM processors wouldn’t be inconsistent with their business or their future strategy.

Why Compete With Yourself?

There is one possible argument against Intel building ARM: it means that Intel would be competing with itself, pitting Atom against its own ARM processor. But plenty of companies compete with themselves all the time, with some success. Toyota has overlapping product between Scion and Toyota. IBM’s POWER and Intel servers compete for essentially the same market. There is plenty of overlap between Boeing products. Yes, GM is an example where competing with yourself didn’t work. I’m not saying that this kind of approach is without risk. But in this case, the benefits may outweigh the risks.

In some ways it is easier to compete with yourself than an outsider, because no matter the outcome, you win. If it turns out that the x86 family just isn’t suited for mobile, or people really don’t want it, Intel is still in the game, rather than playing catch-up from even further behind. And they are in the game with a product people want. Trying to salvage a product that people don’t want to buy is typically futile. IBM nearly died trying to keep the mainframe alive, when it was clear the mainframe was in an irreversible decline. Once IBM realized that they had other product, and that selling something is better than selling nothing, they were able to pull back from the brink.

A Future And A Purpose

Tablets are unlikely take over the entire PC market. They probably won’t take over more than half in the short-term. But consider some numbers. PC sales of all kinds (including tablets) saw almost no growth in 2012. Of the 500 million units moved, nearly 100 million of them were tablets. Clearly some portion of the traditional PC market is being displaced by tablet computers, given the lack of meaningful growth overall. Intel isn’t a big part of the 100 million, so their numbers are already being hit to some degree.

Intel needs to get into the game, and get into it with a product manufacturers actually want. Sure, the Atom processor has all kinds of attraction. But the game appears to be one dominated by the ARM architecture. Intel has the skills and the resources to make ARM processors. I believe it is time they jump in and do it.