The Man Who Thinks Apple Will Buy ARM

It was a balmy evening, thanks to the outdoor gas heaters, on a Californian terrace, and I was talking to the only man I know who thinks Apple will take over ARM.

Sandeep Vij, CEO of MIPS, says he believes in the rumour. That’s probably because he’s about the only person who would benefit from it.

When Apple bought PA Semis, it cut off the company’s external sales.

If Apple bought ARM it would probably do the same thing.

If ARM was owned by a mobile phone manufacturer, no mobile phone manufacturer would want to have an ARM core in his phone.

The whole viability of the ARM business model is that it isn’t Intel’s business model. It allows anyone to buy a licence and make an ARM processor.

ARM’s way means that no one can exert the same sort of control over the mobile phone market as Intel exerts over the PC market.

Even though ARM has a more pervasive monopoly in the mobile phone market than Intel does in the PC market.

This is very frustrating to companies with cores to license, one of which is MIPS.

But if someone buys ARM and either cuts off further development of the ARM architecture, or develops it only for in-house use, then the mobile phone makers would look around for a new core.

And a new vendor of cores could get into the mobile phone market.

MIPS would be the most likely company to take over as the monopoly supplier of cores to the mobile phone market.

Will it?

Well there’s one person who thinks so.


Comments

18 comments

  1. Intel is getting desperate and frustrated as each day goes pass. Intel’s world is crumbling down in front of their eyes, powerless to defend it precious asset which they have dominated with the help of Microsoft for more than two decades.
    CPU from ARM Holding dominate the mobile communication industries and now they are moving up stream to position themselves as a truly global CPU and processor company.
    ARM is moving rapidly to make CPU for the Mobile Internet Devices (MID), and Netbooks. Apple’s iPad is based on the ARM CPU and Nokia has launched a Smartbook based on the ARM processor as well. Traditionally company like Apple, Dell would of naturally chose Intel to power their products but that has changed now, and ARM is replacing Intel gradually.
    Recently the were reported rumours that Microsoft and ARM Holding have signed an agreement which will allow Microsoft to port the new Windows 8 due to be launch in beginning of 2011 on the ARM CPU architecture, if the rumours are to be believed then this can be translated as a disaster for Intel and it feature as an dominant processor maker and certainly will but to the end of WInTel which Microsoft and Intel created some two decades ago and became the de-facto standard.
    source:
    http://www.payasyougomobilephones.mobi

  2. Apple taking over ARM Holding makes no sense to me. ARM holding practices and exercises an unique business model, therefore any technology company will have major problem running ARM.

  3. 1.4 billion devices may have shipped, but how many were based on the original ARM core design and how many on the more complex higher value designs? The ARM7 core license deals were cut in the early days when ARM was breaking into the market. Convincing a licensee with a sweetheart ARM7 core license to upgrade to a more expensive license (ARM 9 for example) is tough. Especially when you can break down your app environment into several ARM7 digestable chunks if you need more processing power.
    Intel developed an ally – Microsoft – whose bloatware made it necessary to upgrade your core on a regular basis.
    ARM does not have such an ally – in fact their whole business model is predicated on keeping core costs down to enable volume.
    So ARM needs another strategy –
    The difference between Intel and ARM? Intel worked to create a market. ARM worked to serve a market. Hence one drives it the other follows it (commercially). And this my friends is the problem with British strategic business thinking and why our US compatriots so regularly take control of markets that we innovated.

  4. The motivation for Apple would be to deny their competitors use of the ARM processor IP, and thus have a commercial advantage over them.
    I think it would be cheaper and more effective for Apple to buy Imagination Technologies instead – as well as the PowerVR GPU and video cores, they would get exclusive access to the communication IP UCCP3 (WiFi, Bluetooth, etc.)

  5. Billy, I’m not pretending to be an expert just trying to start a small debate about why we Brits aren’t very successful at making ‘real’ money out of our invention – even if it is a cliche.
    I suspect there is a little more to ARM than a grad student is capable of and the risks with designing anything complex in-house are bound to be high. Your point about patents is interesting though.

  6. why do you “experts” comapre intel with arm?
    intel sell finished goods (chips).
    arm sell designs – and the customer adds value to the basic core and produces their own finished goods (chips). the wee arm core is a stub of logic in the corner of the chip somewhere.
    remember that designing a risc processor is something a final year grad student does at university.
    the price of the arm design has to be cheaper than developing it in house….. the mountain of software/design tools tethered to the instruction set is the REAL value proposition.
    when do the early arm patents expire? when they pop the clones will arrive. any gate-slinger of even modest ability can churn out a lump of verilog with a arm-compatible ISA.
    up to now, the legal drones at arm have bought-off the odd contender? when the hoards arrive in open-source hardware land – what then? the key is the expiry date of the patents?
    one answer – hard-technology designs – hard-IP so you are providing more of the design block to a customer – maybe allowing the embedding of some secret sauce (low power trannys, clock gating, multi-Vt jolies,etc) – OR hope ya can flog the whole eco-system to a big parent company, sigh and just take a rest – cue Apple?

  7. Robert and Storriefit, I think you have both hit the nail on the head. I know it is a cliché but why are we so good at inventing things and so poor at playing the game (on the edges of legitimacy if that’s what’s required) to squeeze the maximum value from them.
    Competition authorities often just seem like an excuse for people like the EU raising a tax from successful companies. I can’t see it taking Intel down – the US government would invoke a trade war.
    Part of the lack of an ARM netbook story is about Microsoft not having ported Windows – like it or not Linux just isn’t for most people.
    Arguably I guess ARM wouldn’t have got into so many devices if it wasn’t so cheap but does it need to be quite so cheap. Sadly it’s probably too late for them to do anything else.

  8. Spot on Stooriefit.

  9. If we are looking respective performance of Intel and ARM, I wonder how much of the difference might be attributable to Intel’s alledged market dominance abuses vs ARM’s more straight batted approach?
    As David said in a recent post, there is a question mark over whether Intel’s bungs to computer builders have influenced the non-appearance of the ARM based netbook.
    ARM’s approach might actually be the more successful really long term play, if the competition authorities decide Intel’s approach is just not cricket.

  10. IMHO ARM has two problems to address
    – Most TV’s and Video systems use MIPS but most phones use ARM, so convergence in Smartphones and Feature phones is forcing phone designers to decide if the migrate all there MM stuff from MIPS to ARM or all their phone stack and protocol from ARM to MIPS. Any silly decisions by Apple or ARM at the moment would make the decision easy….
    – Second problem : how to capitalize on the huge variety of applications that use ARM. For some apps like GSM phone they MUST settle for less than 5c royalty, for other apps they could easily demand $2 royalty for the same core that is used in the Cell phone. The question is HOW to do this. Unfortunately consumers don’t buy ARM cores directly, so they must sell their IP to design houses and these guys don’t want to pay 5c for one app and $2 for a different app.

  11. Nick – I agree but 3p a chip seems ridiculous – without the ARM chips devices such as the iPhone would likely not exist. The licensing fees are wrapped up in their total turnover.
    Fring – Exactly. ARM have a fabulous technology – from that perspective it is a great story – but they have done a very poor job of getting value out of their innovation. This technology has as much impact as the Intel processors but the cash generated from it is comparatively poor. If you take profit after tax instead of revenues its the same story $5bn on $35bn of turnover v’s £40m on £300m. Interestingly even though Intel has all this manufacturing capacity ARMs net margin is actually worse.

  12. “I don’t think they’ll buy ARM though because it’s too big of an investment.”
    No they won’t buy ARM because they won’t be able to get it past regulators, both European and American. And that same thought applies to most other companies out there. Just the idea of cutting off licensing would be enough.

  13. Well yes Chris, the comparable turnover is rather meagre, but then it’s two different models. ARM is a fab’less chip designer relying on licensing deals, whereas Intel has untold billions tied up in manufacturing and silicon research.
    But I agree, ARM’s turnover seems rather small for the amount of business the do.

  14. Chris – isn’t one of the main reasons that ARM cores are in so many devices is their low price? Raise the price and the volume goes down. There’s also licensing fees as well as royalties to consider in ARM’s income.

  15. I am not if it is really good news for MIPS.
    Yes, short term MIPS can benefits, but it also really means is that the IP Business Model is dead.

  16. Guess Apple most likely decided not to buy MIPS 🙂
    At least one person thinks so.

  17. The increased importance of mobile devices make me a 100% believer that Apple will continue to invest in ARM devices. I don’t think they’ll buy ARM though because it’s too big of an investment. What they will continue to do is buy smaller companies for 300 million or 1 billion here or there to build up there own world-class design team and continue vertically integrating from there.

  18. The thing that amazes me is that ARM cores were put into 1.4bn devices in the first quarter of 2010. Yet they received a meagre £43.2m in royalties (3p per device if I have my sums right). This surely doesn’t represent the real value of this technology. Arguably ARM chips have the same impact as Intel ones – turnover $35.1bn (ARM £305m in 2009).
    From a financial perspective this is not a great British success story!

Leave a Reply

Your email address will not be published. Required fields are marked *

*