The Future of Computing Part 3: Death of the PC

At the end of Part 2 (Part 1 is here) I asserted that a new force would enter the world of computing completely changing the landscape. I stated that Microsoft will lose this battle. I lied, there is more than one force, but one way or another there will be one result.

The PC will die

I expect one thing to happen that the industry will have no defence against: The PC is going to be, for the most part replaced. You may think this a somewhat bold assertion given it’s success, but in order to understand why you must first understand that the economics of the PC market are somewhat different compared to the rest of the world, this alone almost makes the PC’s replacement a forgone conclusion.

When PC and Mac fans duke it out on-line the Mac is commonly criticised as being propriety hardware whereas the PC is made up of cheap commodity hardware. This argument hides a truth rarely revealed: The PC itself is a highly expensive niche product. x86 processors power more than 90% of the desktop market and have a sizeable chunk of the server market. Yet these same processors make up just 3% of the total number of CPUs shipped per year.

The CPU industry as a whole ships more embedded CPUs per year than the number of people on the planet – and they have done this for the last 8 years running. Ubiquitous computing is not coming, it is here already, but no one seems to have noticed. Phones, Cameras, Hi-Fis, TVs, DVD players, MP3 players – they all have CPUs. These use 32 bit processors which are only a minority in the CPU market, the real action [1] is 8 bit CPUs which are the market leader [2], this is liable to inspire most computer enthusiasts with horror but just consider that the embedded industry spent the last decade upgrading from 4 bit CPUs!

Despite it’s relatively tiny size the PC industry gets the majority of the profits due to some very large margins. What’s more, Microsoft have margins which dwarf the computer or semiconductor manufacturers, including 228% in their “InfoWorker” (read Office) division and a whopping 415% profit on Windows, by contrast Dell made 6% (figures for 3rd Quater 2003) [3].

It is these huge margins combined with the under-utilised capabilities of modern microprocessors that mean the x86 and especially Microsoft are vulnerable. It would be very easy to get into a price war with the PC industry – and win.

A second difference from the rest of the world is that PCs need to constantly upgraded. The entire industry is based on obsolescence. With a constant flow of people and businesses replacing computers with new ones Computer, Semiconductor, Operating System and Application vendors have done very nicely thank you over the past 20 years. However computers and software have now become so powerful and feature rich that the endless upgrade cycle is beginning to come to an end. In other industries users only upgrade when the product becomes useless, TVs and other home appliances can be in use for many years before needing to be replaced. This is also starting to become true for the Computer, I have a 800MHz PC and it’s just fine, I could upgrade but the fact is I have no pressing need to.

The challenge that the PC is going to have to face is a computer not based on high performance, high margins or built in obsolescence, it will be a small, ultra-low cost computer.

Return of the Mac

The original idea for the Macintosh was for an ultra low cost computer, it would have been very different had that vision of been followed, It would of had an 8 bit CPU, no mouse and no GUI. The end product was changed drastically after Steve Jobs [4] took over as the project leader and he changed it into the vastly better product later launched.

Ultra low cost computers have been done before, the PC Jr [5] was one example. By today’s standards it’s $1000+ price tag is not exactly low cost but then it was something of a bargain. However users then needed all the power they could get and it subsequently never sold.

A low cost computer is unlikely to be of interest to most computer enthusiasts but that is irrelevant since they only buy a fraction of the computers sold. Most computers go into offices where they do exciting tasks like e-mail, web browsing, word processing, presentations and spreadsheets.
You do not need a 1GHz CPU to do these tasks, in fact most of these tasks have been done on much, much slower computers for many years, a 100MHz processor will do fine, but our low cost system wont be quite that slow.

Someday someone with enough resources is going to start building them, I expect them to arrive from Asian countries most likely running a form of Linux. Education will see the immediate benefit followed by governments, business users will follow when they see the millions saved in large rollouts. Slowly at first then faster they will start eating away at the PC market, they will eventually take over from below. Microsoft will no doubt fight back with bigger, better software with ever more features but is this really important in business? Microsoft will have to have products good enough to justify their colossal margins, with low cost competition they are going to have to fight hard to avoid having to slash prices. The strategy I suggested they are contemplating for PCs in Part 2 will not help them here, they will not be able to get special hardware added to these boxes.

Who will it be?

Who is going to make these ultra-cheap computers? Dell? HP? Apple? Microsoft?
All of these companies are capable of producing the hardware but would they understand what they need to do? The technology industry works on certain assumptions and in order for this to succeed those assumptions have to be set aside. Most people are naturally conservative, businesses, and especially large businesses – even in the technology sector – are especially conservative. Whoever does this is not going to be a conservative, and most likely wont even be from the technology industry.

The business model for this is completely different than anything in the computer market today, this business will operate in a manner similar to that of a cheap TV vendor, nothing like a computer vendor. Whoever sells these machines will have to understand that the users will buy them and use them for years, that they wont get sales from obsolete systems every other year and high margins are a big no-no. This product will make it’s money purely by volume, just like normal consumer products.

That’s not to say it’ll be easy just if you know that the rules are different.
The hardware can be standard embedded parts or perhaps a custom design (which will save costs when produced in large quantities). Before even starting production you need to find an OS which can run applications acceptably on a machine which is lower spec than any PC on the market today. Programmers today are used to multiple GHz, hundreds of Megabytes of RAM and masses of Hard Disc space so there’s not that many OSs about these days from which you can chose. Even then you’ll need high volumes from the very beginning and that costs money even for low cost hardware, got $25 million spare? Then you have to sell them, but that’s a whole different story.

Giants fall

In twenty years time there is a distinct that Microsoft may be just a bit player – if they still exist at all. If it sounds off the wall that a small computer could unseat the biggest giant, remember that in Part 1 I stated that we can look to the past to predict the future, this sort of change has happened before in other industries.

I am predicting a change which already has historical precedent – even in the computer industry. Go back 30 years and only one of the four companies I listed above actually existed.
The PC itself brought in a massive change in the computer industry, turning it on it’s head and removing it’s biggest player from the top spot, IBM used to be the biggest company in the industry and indeed it was they who produced the “IBM PC”, the ancestor to today’s PC, but it was that same PC which went on to topple them from the top and relegated them to producing expensive business systems and chips (that said they do very nicely doing those and are by no means small).

I believe this industry is about to undergo massive changes, the rules and assumptions which have kept this industry in it’s current shape for the past twenty years are changing, Microsoft will not have 415% profit margins for much longer. Nobody knows what’s going to happen, we won’t know until it’s over, by then it will be too late and the new powers will have taken the reigns.

Rumors of My Death Are Greatly Exaggerated

The PC will almost certainly not “die” as such, these small computers will take over much of the market but not all of it. There will still be a market for high end performance PCs but it will be vastly smaller then the current PC market. Without the price advantages PC vendors will have to resort to new tricks to sell their wares, this will be difficult on Dell and HP but wont be a big threat to Apple who already concentrate on users who generally need more power than the low end system will provide.

It’ll all happen when someone produces a computer and sells it for $100. Yes, it’ll be that cheap and it’ll include the OS and all the Applications necessary to be useful. I expect the price to fall significantly later but with sufficient volumes a $100 (retail price) computer can be produced today. There are choices for the OS but it could obviously not be Windows, it costs more (outside the US) than the entire system!

The Second front

The PC is under threat from other areas as well, it’s not obvious right now but advances in other existing areas will produce new competitors which may provide an even greater challenge to the PC.

The PC’s little brother, the PDA is also under threat, the stand alone PDA will not be around for much longer. The foe is sweeping aside the PDA with ease, but it is this same foe which will later go on to threaten the PC as well.

The PDA is under threat from Phones which are now becoming increasingly powerful, SmartPhones are now outselling PDAs 2 to 1, I expect this trend to continue. PDAs will continue to lose market share until it pretty much gets all taken away by the SmartPhones.

This will be especially prevalent In Europe where Phones are generally sold with contracts. You pay the contract monthly and a big chunk of the price is knocked off, in many cases you can expect to get the phone for no cost whatsoever. No PDA vendor can compete with free so expect even the lowest end PDAs to become under pressure, as a result they too will get telephony functionality built in.

Eventually I expect the stand alone PDA to die out as a consumer product altogether but it’ll probably continue as a industrial product in some form.

SmartPhone Operating Systems

A new OS battle is opening up on SmartPhones. Here we’ll see a fight between Symbian, Microsoft, PalmOS and Linux:

Microsoft will keep pounding away but without the market’s trust will have difficulty getting any of the large vendors. Expect to see more no-name and Telco branded devices from Microsoft. The only exception to this is Motorola but they are also working on Linux based phones.

I don’t think Linux will get very far though it could do well in low cost devices. As with other markets the Linux kernel may look like an attractive option because of it’s “no cost” but the surrounding software will remain firmly closed source and will have to be paid for. Consequently the price advantage is not all it’s cracked up to be and with vendors heavily customising devices a complete solution with a good track record such as Symbian will continue to look like a better option.

PalmOS is a bit of an unknown at this point for phones, especially their new OS 6, which while no doubt technically good (coming from the BeOS team) is still new. That said I expect the usual PalmOS vendors will use OS 6 in new PDAs and SmartPhones though even Palm themselves are said to be considering using other Operating Systems in future devices.

Symbian has the support of the all the major vendors but that could change. Symbian was technically very good before they even started working on Phones in 1998. At this point they are providing a complete solution to Phone vendors and have many new Phones in the works. Nokia is also producing a solution for Phone vendors based on Symbian.

With Nokia now in charge of Symbian [6] things could get complicated, the Phone vendors will not want Microsoft replaced with another Microsoft. This could bode well for the competitors and I expect if the vendors are unhappy with Nokia’s control they could go running to the other options. This could do PalmOS 6 a world of good. On the other hand Nokia was said to largely be running the show at Symbian for some time anyway so it may not make much difference.

OSs aside the software market for Phones is starting to take off and I expect this will continue, Java is especially popular as it works across different Phones and OSs.

One thing I do not expect but I’d like to see is a return of Psion to the PDA domain, they never made anything that could be considered even vaguely Phone-like and with all the PDAs turning into phones I expect many PDAs will vanish, I can see their product fit into a neat niche – but even it would have to include a Phone. I do not expect this one little bit but perhaps someone else will see this opportunity.

To 3G or not to 3G

The day of 3G Phones seems be taking awfully long time to get here. That said that day is coming and the devices are beginning to trickle out.

Some suggest that 3G will be a flop, that 802.11 (aka WiFi) will take over instead.
I don’t see that myself, there are a lot more Phone users than Laptop / high end PDA users. 802.11 has limitations in that it’s target market is itself limited. It’s also limited in it’s range, an area where any phone will have an advantage.

I don’t see one technology killing the other, I expect one to live alongside the other but I expect 3G phones will dwarf 802.11 usage once 3G begins to take off. That said Phones will include 802.11 at some point so everyone’s going to have it anyway, this could lead to a price war between Fixed line and mobile Telco’s driving the price of both fixed lines and 3G down – at least in areas where the two overlap.

3G is still stupidly expensive and will remain little used until it becomes economical for normal people to use it. Once it does it’ll allow high resolution images to be sent and we can expect picture messaging to take off even more. I’m not so sure about video but I do expect it to catch on to a degree although perhaps not as much as the Telco’s would like. It’s apparently quite akward but among others I can see a use for video when you are calling distant relatives.

The Phone’s next victim: The PC

Phones have had Internet capability for some years, it did not take off as much as expected but seems to be catching on now. Once 3G SmartPhones appear I expect Internet surfing over phones to really take off. This will not only take yet more market away from PDAs but also threaten the PC as the centre of their users world. Who’s going to go and switch on their PC if they can browse the web while vegging out on the couch?

This may not exactly sound like a big threat but being on-line is one of the largest uses of PCs. Phones will be getting Hard Discs in the not too distant future and then will become a more serious threat to the PC, for many a Phone will able to store their data and do much of what a PC is used for.

The only thing holding it back will be the screen, but even that may not be a problem for long, put a connection into the base station for a screen or TV along with keyboard and mouse and you’ll effectively have a PC. Eventually you’ll not even have to plug it in, looks like “Ultra Wide Band” will cover that. The same technology will put a swift end to the “media players”, but then I don’t expect them to catch on in any case – who wants to watch movies on a 3 inch screen?

Scanning, Printing and other peripherals will also be plugged into a base station with USB / Firewire or whatever else becomes popular, perhaps that’ll be 802.11’s job. For many there will be no point having a PC at all, you will be able to do everything on a Phone.

Volumes? No problem, Mobile Phones are produced in volumes some 4 times higher than the PC. The No 5 mobile manufacturer (LG) makes more phones than the No 1 PC maker (Dell) make PCs.

The No 1 (Nokia) alone make more phones per year than all the PCs (companies and users) combined[7].

For business a $100 PC will be the draw, for consumers a SuperPhone will do it. The PC is under threat from changing economics, lower cost embedded parts and more advanced, more convenient technologies, and I haven’t even mentioned home entertainment boxes…

On the home Front

I have already written my thoughts on TV-PC convergence [8] in which I argued that while the capabilities will be similar the TV and PC will remain separate due to the different ways in which they are used.

When it comes to a fight however I can’t see a PC being a popular replacement for a TV. If the capabilities of the converged TV are sufficient we could see home PCs not being replaced for the more casual PC users. There are no end of PCs out there who’s capabilities are not even touched upon, or in many cases not used at all. There are many computers out there which are hardly ever used, move what it is used for into a TV or elsewhere and the PC has no more reason to exist. For these users the mythical convergence may actually happen and the TV will completely replace the PC.

The death of the PC may be inconceivable for the average geek but then they don’t buy most PCs, PCs are a mass market consumer and business product. Make something better / cheaper / more convenient and the PC becomes surplus to requirements for big sections of that mass market. When that begins to happen the PC will begin it’s downward spiral. The PC will start moving back to it’s original user base – the geeks. The lower volume of production will start sending prices up and this will make the PCs competitors even more attractive.

This isn’t going to happen overnight, but slowly and surly I do expect the pressures to build and the PC will start falling. The PC vendors already know that the endless upgrade cycle is coming to an end and are making a pitch for the living room, looks like it’s gonna be tough[9] and that’s just one of the battles the PC is going to have to fight.

In Twenty years

Fast forward Twenty years and the Computing world will be a very different place, the major players will likely be different and one way or another Microsoft’s absolute dominance will be long gone.

New practices in development, new CPUs, new PCs and new competitors for them.
The computing industry is entering a phase in which it will be consolidating the progress made into a solid infrastructure industry. Simultaneously the technology that made this possible will be rapidly evolving as the computing landscape changes into something we’ve never seen before, the PC was part of that journey, it is not the destination.

There will be winners and losers, most of all there will be change. A golden age of innovation is coming as Phones, TVs and PCs clash for our attention and companies race to produce the biggest selling products. It’s best not to bet on the winners now, the cheapest, most feature rich product does not guarantee anything, the iPod has sold well despite being relatively expensive and it’s not stuffed with features. I expect one of the most important features will be the interface, techies love complex stuff but most users don’t. The company who gets that power to the majority through a good interface will have it made.

PCs will still be around in a somewhat limited form but they will be vastly more powerful. New display technologies will also be around so what remains of the PC will provide a very different computing experience. When we have a Teraflop on the desktop, How will we use these computers? What will we use them for?

In Part 4 I shall go beyond guessing the future of today’s technologies and shall dive into the blue sky of real future computing.



[1] Zilog have just released a new OS for the 8 bit Z80 CPU

[2] Discussion on upgrading from to 32 bit from 8 bit embedded processors.

[3] Microsoft make a LOT of money.

[4] The Mac changed considerably from the initial ideas.


[5] IBM’s low cost ($1000!) PC.

[6] Nokia get control of Symbian

[7] The mobile phone market is rather large.

[8] TV & PC Convergence.

[9] Who do you want your media server from?

Copyright (c) Nicholas Blachford February 2004

This series is about the future and as such is nothing more than informed speculation on my part. I suggest future possibilities and actions which companies may take but this does not mean that they will take them or are even considering them.


  1. 2004-02-11 9:29 pm
  2. 2004-02-11 9:36 pm
  3. 2004-02-11 9:55 pm
  4. 2004-02-11 9:56 pm
  5. 2004-02-11 9:59 pm
  6. 2004-02-11 9:59 pm
  7. 2004-02-11 10:01 pm
  8. 2004-02-11 10:01 pm
  9. 2004-02-11 10:03 pm
  10. 2004-02-11 10:03 pm
  11. 2004-02-11 10:09 pm
  12. 2004-02-11 10:09 pm
  13. 2004-02-11 10:12 pm
  14. 2004-02-11 10:13 pm
  15. 2004-02-11 10:23 pm
  16. 2004-02-11 10:23 pm
  17. 2004-02-11 10:26 pm
  18. 2004-02-11 10:31 pm
  19. 2004-02-11 10:42 pm
  20. 2004-02-11 10:46 pm
  21. 2004-02-11 10:57 pm
  22. 2004-02-11 10:59 pm
  23. 2004-02-11 11:06 pm
  24. 2004-02-11 11:11 pm
  25. 2004-02-11 11:12 pm
  26. 2004-02-11 11:14 pm
  27. 2004-02-11 11:15 pm
  28. 2004-02-11 11:20 pm
  29. 2004-02-11 11:27 pm
  30. 2004-02-11 11:33 pm
  31. 2004-02-11 11:36 pm
  32. 2004-02-11 11:39 pm
  33. 2004-02-11 11:45 pm
  34. 2004-02-11 11:46 pm
  35. 2004-02-11 11:49 pm
  36. 2004-02-11 11:54 pm
  37. 2004-02-11 11:55 pm
  38. 2004-02-11 11:59 pm
  39. 2004-02-12 12:07 am
  40. 2004-02-12 12:09 am
  41. 2004-02-12 12:23 am
  42. 2004-02-12 12:24 am
  43. 2004-02-12 12:29 am
  44. 2004-02-12 12:37 am
  45. 2004-02-12 12:46 am
  46. 2004-02-12 12:59 am
  47. 2004-02-12 1:25 am
  48. 2004-02-12 1:36 am
  49. 2004-02-12 1:39 am
  50. 2004-02-12 1:48 am
  51. 2004-02-12 2:07 am
  52. 2004-02-12 2:30 am
  53. 2004-02-12 2:37 am
  54. 2004-02-12 2:37 am
  55. 2004-02-12 2:41 am
  56. 2004-02-12 2:44 am
  57. 2004-02-12 2:54 am
  58. 2004-02-12 2:57 am
  59. 2004-02-12 2:58 am
  60. 2004-02-12 3:25 am
  61. 2004-02-12 3:28 am
  62. 2004-02-12 3:40 am
  63. 2004-02-12 4:22 am
  64. 2004-02-12 4:29 am
  65. 2004-02-12 4:38 am
  66. 2004-02-12 4:46 am
  67. 2004-02-12 5:25 am
  68. 2004-02-12 5:50 am
  69. 2004-02-12 7:11 am
  70. 2004-02-12 7:25 am
  71. 2004-02-12 8:05 am
  72. 2004-02-12 8:10 am
  73. 2004-02-12 9:35 am
  74. 2004-02-12 9:46 am
  75. 2004-02-12 10:43 am
  76. 2004-02-12 11:00 am
  77. 2004-02-12 12:21 pm
  78. 2004-02-12 12:36 pm
  79. 2004-02-12 12:48 pm
  80. 2004-02-12 1:54 pm
  81. 2004-02-12 4:15 pm
  82. 2004-02-12 5:06 pm
  83. 2004-02-12 7:39 pm
  84. 2004-02-12 8:52 pm
  85. 2004-02-12 10:07 pm
  86. 2004-02-12 10:55 pm
  87. 2004-02-12 11:04 pm
  88. 2004-02-12 11:54 pm
  89. 2004-02-13 12:52 pm
  90. 2004-02-13 1:02 pm