x86PC Has Peaked. Here’s Proof.

One picture is worth millions of words. Here’s a site that has plotted sales/shipments of various platforms of personal computing over the decades. The PC has obviously peaked while new technology climbs like a scalded cat. It’s also clear that the new technology is just getting started… and Android/Linux is overtaking whatever Apple puts out.

There are many reasons for why the new technology performs this way:

  • the old technology is bulky, expensive, heavy, hot and noisey,
  • the new technology is small, cheap, light, cool and quiet,
  • the new technology is easier/faster to bring to market mostly because it is cheaper and has so many advantages,
  • more choices stirs interest in consumers, and
  • more options stirs innovation in producers.

The counter argument is that the x86 PC can do more faster but that is irrelevant as consumers don’t want/need to do more. “More” was just marketing hype from Wintel over the years to bring on new sales rather than any kind of necessity. A lot of basic roles PCs performed have largely been taken over by servers and USB devices: storage, computing, manipulation of data, and of course servers do a lot of stuff PCs cannot, like the social media. PCs do not scale well enough to do that.

The only question in my mind is whether or not M$ can swing the monopoly it had on x86 PCs to ARM. I think there’s not much chance of that working because the new technology is all about small and cheap whereas Wintel is all about big and expensive. Some things need to be big to perform well, like vaccuum cleaners, but a personal computer is not one of them. Moore’s Law, networking and ARM have eliminated the need to be big in IT. One does need to be flexible and the EULA of M$ and the need to cool x86 make Wintel deprecated. Intel is trying to remain relevant by it’s Atom line and pushing GNU/Linux but Atoms need to be one or two steps ahead of ARM to be in the same league of energy consumption. Eventually Atom and ARM will both be so low in energy consumption that difference may not matter but that is still a year or two away and Atoms still cost far more than ARMed CPUs to produce. There may continue to be a need for desktop/notebook personal computers but there is less need for them to be run by Wintel. The same is true for “social” servers.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology. Bookmark the permalink.

14 Responses to x86PC Has Peaked. Here’s Proof.

  1. Clarence Moon says:

    I don’t think that you have the right idea about how Google collects money via clicks, Mr. Pogson. People and companies pay Google for hits on the person or company’s own site. Just going to Google doesn’t pay them a cent. The person has to link to the advertisr’s site for the “ka-ching!” to occur.

    And I do use Chrome on one of my machines, Mr. Pogson. It does have a slightly different feel than IE, I will admit, but I don’t know if it is better.

  2. Clarence Moon wrote, “I don’t see where there would be more clicks because of Android.”

    Obviously, you have not used Chrome browser. Anything typed into the address bar gets sent to Google by default. A user would have to exert himself to change that behaviour. Enough do not that the strategy of bundling the browser with the OS pays handsomely.

  3. oe says:

    “Future considerations” —> “Repeat Business” —-> “Loyal Customers”….perhaps Google isn’t so dumb by giving up some today for more tomorrow. Apparently they have faith in the quality of their products as they make it REALLY easy to get out of their Docs/+/Gmail with data export you’d never see in the likes of Azure/Facebook/Hotmail/Ebay etc.

  4. Clarence Moon says:

    I rather think that all browsers lead to Google, Mr. Pogson. Certainly mine does. I think that the number of clicks depends on the aggregate interest of the users and they click when they are curious about something. I don’t see where there would be more clicks because of Android.

  5. Clarence Moon wrote, “Google still does not make out since a click-through from an Android phone is no more or less valuable than a click through from a Windows computer.”

    Ah! You forget that people use Android/Linux a lot and there may well be more clicks from a user of Android/Linux than that other OS. It’s all in the browser, you see, which is much more likely to be found on a Android/Linux machine than on one with that other OS. You see, bundling works… M$ proved that and Google is exploiting the same phenomenon.

  6. Dr Loser wrote, “don’t forget to publish the quarterly M$ results tomorrow or on Saturday”.

    Done! and you don’t have to wait.

    Revenue for the client division was down 6% and income down 11% as expected while PC shipments were flat. M$ is no longer tracking PC shipments because people aren’t buying their stuff. Now the shortage of hard drives will kick in and retailers will be looking for something else to sell and voila! GNU/Linux will appear on shelves along with Android/Linux. People want small cheap computers and not Wintel.

  7. Dr Loser says:

    And don’t forget to publish the quarterly M$ results tomorrow or on Saturday, Pog. You’d only let me down if you didn’t. I’ve spent a lot of personal capital explaining to people who think that you’re a rural nutcase that you are, in fact, a decent, honourable, and above all truthful man.

  8. Dr Loser says:


    “Future considerations,” you say? Wouldn’t that be awfully close to Unearned Revenue according to you?

    In deference to Clarence, I should point out that it’s closer to what he would regard as Goodwill.

    But, let’s be honest, it’s bull either way.

    “I don’t know how much developing Android costs but it likely is quite small compared to all Google’s expenditures because they are mostly re-using FLOSS and other people write the apps.”

    Ah, a false comparator. Of course it likely is quite small compared to all Google’s expenditures. They could write the goddamn stuff in gold leaf on rare palimpsests through the agency of resuscitated Medieval Monks and it would still “likely be quite small compared to …”

    I’m thinking of writing Eric a letter on this one. “I already know how to re-use FLOSS. I am an industry expert of 20 years. I understand that you have just burned something like a billion dollars on doing something that I, and a couple of my eye-balling friends, could do for a scant $50 million. I think this represents fair value; and, if you disagree, I will take my talents and several highly-connected eyeballs to my local VC and build an equally capable competitor.”

    No worries on this one, Robert. If I get the signoff on $50 million, I will personally ensure that you have a free supply of dehydrated potatoes for life.

  9. Clarence Moon says:

    More to the point, though, Microsoft’s business is selling licenses to use their software. Many things affect that business, of course, and Google’s giving Android away likely has some effect on how much money people are going to be willing to pay for how many Windows licenses. It is not a one to one sort of thing, though. Someone using an Android phone is not necessarily any loss to Microsoft in terms of Windows license fees.

    For that matter, if you were at all correct, Google still does not make out since a click-through from an Android phone is no more or less valuable than a click through from a Windows computer. Google gets the revenue either way and gets no more by substituting an Android click for a Windows click.

  10. Clarence Moon says:

    Well, it is still a matter of opinion, Mr. Pogson. On a more practical note, Google took it in the shorts today after the market closed and they announced results below expectations. Their revenue per click was off some 8% and their costs per increasing clicks was higher than before, too. Maybe they are running out of room?

    Microsoft did suffer some reductions in Windows client sales, attributed to few PCs sold due to various bottlenecks, mostly the hard drive shortage. But they more than made up the shortages in other areas, notably servers, office software, and Xbox software. They’re not going away as soon as you would have it, Mr. Pogson.

  11. Clarence Moon wrote, “Google got approximately butkus for Android.”

    You are comparing apples and oranges. Google gets tons of advertising and app revenue from Android and I would bet Google gets a lot of “future considerations”. 96% of their revenue is from advertising so you could say the same about Gmail, and Google+ and search being freebies that do not generate revenue directly but indirectly those things are fire-hoses pumping money to Google. Android/Linux is the same sort of thing. Google gets more ad revenue because they provide the software platform people are using, mostly through the accompanying browser. Chrome browser is a huge asset for Google and the more copies running out there the better.

    I don’t know how much developing Android costs but it likely is quite small compared to all Google’s expenditures because they are mostly re-using FLOSS and other people write the apps.

  12. Clarence Moon says:

    Well, Microsoft got over $20B for Windows client licenses alone in 2011, Mr. Pogson, while Google got approximately butkus for Android. Analyst estimates put them around a billion bucks in the hole when expenses were totaled up, making their bubble more like a sinkhole. When it comes to blowing bubbles, you need to follow the green, eh?

    How much did Apple get for iOS? About the same, I think. Any others?

  13. Clarence Moon wrote, “Such an individual bubble diagram continues to show Windows PC OS as the biggest software game in town.”

    Uhhh no, it does not.

    M$’s client division revenues in 2011 were mostly about sales of 50 million licences for that other OS per quarter. Meanwhile Android/Linux was on hundreds of millions of smart thingies, iOS was on hundreds of millions more and GNU/Linux was on tens of millions of devices shipped. MacOS was on 16 million units. So, M$ is being swamped by client OS shipping from other suppliers. Those non-M$ OS have some value per unit. Even at $10-$20, M$ is behind. Worse, for M$, is that there is nothing to prevent these other OS from gaining share because M$ cannot bully the OEMs as they did in the past.

    Clarence’s argument is based on the circular definition of a PC as a PC with that other OS. In reality more computing was done in 2011 on client machines using other non-M$ OS.

  14. Clarence Moon says:

    I need to note that the author of this tortured analysis seems to have found none of the 40% of non-Windows PCs that you believe are being manufactured annually, Mr. Pogson. Also, it is perhaps being overly redundant to note that the decline, if there is one, in PC volume share is due to including a lot of things that are not particularly fungible with PCs. Why not add iPods and routers using embedded microprocessors to the litany? That would show an even earlier and deeper decline in PC volume relative to everything else.

    For any practical matter, it is better to treat each of these phenomena as individual markets and see what financial affect they might have on the world’s economy. Such an individual bubble diagram continues to show Windows PC OS as the biggest software game in town.

Leave a Reply