steve's blog

thoughts on technology, business and strategy

Search

How Intel Screwed Itself Out Of The iPad Revolution

Image001

Intel made a couple of strategic gaffes along the way to being left out of the Apple iPad:

(1) exiting the ARM-based mobile CPU business entirely in 2006 (pre-iPhone)

(2) assuming that PC technology would continue to extend lower & lower into mobile devices

Here is a brief history:

2005: Mini coup for Intel - Apple announces it will abandon PowerPC and adopts Intel across its entire MacBook line. This will go down as the apex of Intel’s reign in computing, as x86 processors own almost every server, desktop, and laptop.

2006: Intel exhibits a level of hubris that is shocking in hindsight—they exit the mobile phone market entirely, selling off a successful ARM processor business (called XScale) to Marvell Semiconductor. XScale had mindshare, scale and design wins, but Intel gave up. Why focus on these PDA/toys when we can win computers?

2007: The iPhone dawns a new age in mobile computing.  Following its startling success, Intel restarts a ‘low-power’ (mobile) device strategy, only this time they take laptop technology (x86), strip it down, and design ‘Atom’. Atom has since come to dominate Netbooks. But Atom has zero market share in cell phones. Once again for emphasis. Zero.

It’s become clear in the last few days that the iPad is a new class of devices. Not the tablet-of-old pushed by Microsoft, but a truly disruptive era of computing, one that bypasses Intel entirely. Intel simply couldn’t fathom that an iPhone class of computers would disrupt PCs from the bottom up, which is exactly what happened.

And it’s not just the iPad. Google Chrome OS tablets won’t run Intel either, because the divide between ARM and Intel is not narrowing. In fact, Netbooks may even ditch Atom in favor of ARM processors (Windows bloat may prevent this). Google wants this to happen with its cloud initiatives running on Chrome OS.

Little known secret - Intel has a dedicated team of people which caters to Apple full time. A private team of Intel folks firewalled from the rest of Intel (per Steve Job’s demands for total secrecy). Not only sales, but also product folks. You can bet Intel’s Atom CPU roadmap was pitched to Apple for tablet-type devices from day 1.

Intel is fighting back using massive R&D dollars to get back in. However, even though Intel Atom CPUs use more advanced process technology (45nm), they lag ARM processors (the Apple A4, nVidia Tegra 2, Qualcomm Snapdragon etc) from less advanced fabs in the area that matters most - performance per watt of power consumption. Simply speaking, this is performance for a given battery life.

Losing the iPad is symbolic of losing tomorrow’s computer, and it looks a lot like mobile computing will pass Intel by in classic disruptive fashion. Tablet computing - as dawned by the iPad - is the first real threat to Intel's hegemony, and you can bet they know that...

 

12 comments
Apr 05, 2010
Mark Essel said...
Whoa.
Excellent write up, appreciate it.
Apr 05, 2010
steve cheney said...
thanks a bunch mark
Apr 07, 2010
 said...
Good write up. Except for the part "Intel exhibits a level of hubris that is shocking in hindsight ... but Intel gave up", which exhibits hindsight bias, as already admitted! Intel made a bold bet (just like moving from memory to processors) which did not succeed. Taking risks necessarily involves making bad calls - being "shocked" at "hubris" and "giving up" is either naive or sensationalism.
Apr 07, 2010
steve cheney said...
Glad you liked it. if arrogance is not the right word, maybe extremely poor judgment is. keep in mind intel made a conscious choice to walk from the mobile market. This can't be over-stated.i don't think it was balancing risk taking - they just missed out on the market.
Apr 07, 2010
Sironfoot said...
Intel has been caught with it's pants down before. Remember the Prescott Pentium 4's, bad overal architecture leading to massive power consumption & poor performance versus competition, basically it was the Windows Vista of CPUs. This allowed AMD to totally sale past them in both price and performance. Then Intel unleash a radically different architecture in their Core series, and completely turned their fortunes around.

No doubt they've seen the writing on the wall, and have a big enough R&D budget to turn things around for them in the mobile space.

Apr 07, 2010
steve cheney said...
good comment - yes intel is a master at reigning in any breakaways by competitors. But it ain't AMD this time, and PCs are a completely different sales dynamic than smartphones. Intel is so far behind ARM in mobile it will take many years, if they can do it at all
Apr 13, 2010
steveringer said...
Steve,

This is a good, thought provoking blog, but I have to take issue with your conclusion that Intel made a strategic blunder in abandoning the ARM market. In fact, I think you go it exactly wrong!

It’s true that losing the iPad design win is a big blow to Intel and due to a strategic mis-step. But their error was in waiting too long to abandon ARM in favor of creating a low-power Intel Architecture product line (Atom). Intel had absolutely no strategic competitive advantage in the ARM market, so their best hope there was to be one of many minor players in a market with low margins. This would have siphoned resources and focus away from their main IA designs while at the same time helping ARM to cannibalize Intel’s main (high profit) products from the bottom up.

By developing a low-power Intel architecture based product line, they leveraged the strategic competitive advantage of the massive installed base of apps that run on IA (not to mention 30 years of IA manufacturing experience). This has already allowed Intel to create the Netbook market segment which showed rapid growth during the worst economic downturn in 75 years. It’s also allowed them to take the fight to ARM rather than wait for ARM to slowly devour them from below.

Intel’s mistake was not making this move earlier. Remember that Atom is still on its first generation of process technology. It was launched on Intel’s 45nm process and won’t become competitive with ARM in smart phones (and tablets) until it migrates to 32nm in 2011. If Intel had made the move to low-power IA two years earlier, they would be a serious player in the Smart Phone processor business today. Nevertheless, they’re far better off than they would be had they stayed as a bit-player in the ARM market.

Apr 13, 2010
steve cheney said...
Thanks for the kind words on the blog and for visiting.

I don’t think we are disagreeing on most points.

Intel doesn’t like non-PC markets because they lose their incumbent advantage. I agree that anything Intel would have stuck to a few years earlier could have gotten traction. Xscale had designs in Palm.

From a pure strategy standpoint, it seems what you are suggesting, is that it doesn’t serve Intel to make ARM processors because they are not proprietary. I agree with that wholeheartedly

But just as Intel has never had success in low end embedded designs with x86, they weren’t destined to do well in mobile compute, no matter when they would have started.

This is perhaps for a different post…but Intel has failed BADLY in communications and embedded markets because they don’t understand how to sell into these. Their failures here are very well documented, they spent $2B in comms acquisitions during the boom and killed about 75% of these product lines.

Because smartphones were not like PCs, Intel could not figure out how to sell when competing against companies like Infineon, TI etc.

Lastly their netbook dominance is totally superficial—this is an artifact of windows PCs, and no one will take them out there

Apr 13, 2010
steveringer said...
Mr. Cheney,
Good points! Intel's comms acquisitions were indeed spectacular failures. It seems the real challenge for them is to suceed outside of the WIndows world. There's existence proof with Mac OS and Linux, but embedded and smartphones will be very challenging. What ever happened with Intel's acquisition of Wind River? Wasn't that supposed to be their ticket into the embedded space?
Apr 14, 2010
steve cheney said...
i agree, windriver will take time to show - focus is probably multicore / parallel stuff
Jul 26, 2011
worldpharma said...
It's always my pleasure to read this type of stuff.I am very much interested in these types of topics and it's my habit to read this.
Generic Viagra
Jul 26, 2011
Kevin Anchi said...
Thank you for sharing to us.there are many person searching about that now they will find enough resources by your post.I would like to join your blog anyway so please continue sharing with us.
Free Credit Score

Leave a comment...

steve cheney

steve cheney

Engineer with an MBA.

Current entrepreneur.

Former programmer, marketer, investment banker, and vc.

My Other Sites
 
Subscriptions

Contact me
Follow me on Twitter