The coming ARM vs. Intel PC battle

Posted by By Rob Enderle, Principal Analyst, Enderle Group

Analyst Opinion – I am seeing a lot of ARM activity at the moment that appears to be drifting towards the x86 space. ARM is the most common processor used in smartphones like the iPhone.  And I’ve been waiting for a major push by ARM into the netbook space currently dominated by X86.  The first visible broadside happened this week when Marvell announced they were initially going after the home storage space with low cost ARM based storage appliances, but the goal appeared to eventually be PCs.  

However I've been hearing a lot of activity surrounding Google's netbook and how it may actually use ARM and not X86 to get to its very aggressive target price point of under $300 before carrier subsidy (the subsidy will likely make it free in a number of markets, if the price point holds).
 
It appears that others are quietly working on ARM-based netbooks behind the scenes. While much of the really recent ARM netbook activity surrounds Qualcomm's ARM Snapdragon package, which apparently has a number of PC and cellphone OEMs really excited at the moment, ARM in general appears to be stepping up to a fight with x86 and Intel. This should be an interesting environment to watch, partially because it has Microsoft implications as well.   


ARM vs. x86 and Microsoft

Intel started this fight by challenging ARM with its Atom processor, which is moving downmarket and towards smartphones.   Apparently, the major ARM vendors are feeling the threat, are now moving upmarket and are beginning to make their run at low-end PCs and storage appliances to put the pressure back on Intel.
    
Microsoft is largely an x86 vendor, which means that most of these ARM implementations don't run a version of Windows, but run a version of Linux instead. What ARM brings to the table is very low power requirements for a given level of processor performance.  Where the ARM platform historically falls short is in multitasking, software breadth, and consistency between versions. Especially this last one, where x86 remains relatively constant so that a piece of software can generally run on x86 products over a decade old (with some OS-based exceptions) each version of ARM generally requires a platform rewrite and this will make software offerings like application stores rather interesting to manage over the long term. Hardware virtualization, at least for the short term, will be difficult, because there isn't a lot of extra performance overhead to run a virtual machine on ARM yet.
 
This has traditionally made ARM better on products that are closer to appliances like cellphones and x86 better for PCs. Storage devices like what Marvell is anticipating fall closer to the sweet spot for ARM, while netbooks, which are currently more like notebooks, do not. At least not yet, but that could be changing.
 

The cloud:  The great leveler

What may make the big difference is the concept of "The Cloud" or a set of services that reside on the web that make these devices work. The more software that doesn't run locally there is, the more problems represented by ARM are mitigated and the more important connectivity is. You start tacking on appliances like the storage appliances Marvell is imagining and you have the potential for an ARM-based ecosystem that could grow and challenge the currently dominant PC ecosystem.   This new ecosystem could have the benefit of both low cost and low energy use.
 
Still, and at this writing, ARM doesn't do things like full Flash well and has technology constraints not enjoyed by X86 and while the part remains more energy efficient, much of the power for a product like a netbook goes to the displays and hard drives, which have nothing to do with the processor.
 

Intel’s exposure


Intel and its aggressive Atom placement is strong enough to hold ARM off, but they are currently trying to drive the netbook space back up to their Core2 level technology, which is much more costly than Atom is.  The result is a bigger price gap between Intel and the challenging ARM product.  

If ARM can gain a foothold before Intel can react, it will have a defensible beachhead and Intel will have grown a niche competitor into a primary competitor that has aspirations across most of Intel’s desktop offerings and a platform that may be good enough in a cloud-based world.

   
Wrapping up  

What is creating at least some of the opportunity for ARM is Intel's pricing policy for Atom, which is seen as too restrictive by OEMs and is forcing them to seek alternatives. AMD has an Intel mobile based competitor coming as well, though theirs is x86 based. If ARM can get a foothold in the netbook segment before Intel can take significant share of smartphones, we will see an interesting battle.  

If Intel moves down market into cellphones before ARM has a defensible position in netbooks and appliances, then ARM will be reduced to a note history books and will get purged from the market.  It will take a number of years for this to all play out and it is largely dependent on how well Intel executes on the next generation of Atom, how aggressively Intel prices and restricts the resulting solution, and how good offerings like Snapdragon actually are.

However, it is undeniable that end result for us, regardless of the outcome of this fight, will be less expensive as well as thinner and lighter notebooks that are tied to web based services and will be most likely subsidized.

That could be a good thing depending on how well those services turn out.  We will see who gets this right first.     


Rob Enderle is one of the last Inquiry Analysts. Inquiry Analysts are paid to stay up to date on current events and identify trends and either explain the trends or make suggestions, tactical and strategic, on how to best take advantage of them. Currently he provides his services to most of the major technology and media companies.

The opinions expressed in this commentary are solely those of the writer.