Post-PC era means mass extinction for personal computer OEMs

Post-PC era means mass extinction for personal computer OEMs

Summary: HP is laying off more than 27,000 employees and Dell's Q1 2012 earnings were weak across the board. What does this mean for the future of Personal Computer OEMs?

This article is an expansion of Jason's arguments from the ZDNet Great Debate: Can PC makers survive in a Post PC world?

Just over eight months ago, my colleague Zack Whittaker and I were were both standing on our virtual podiums debating whether "Post-PC" was actually real or if it was bunk. In the end, while Zack put up a good fight that the PC would never die, the arguments favored that the Post-PC world was truly upon us.

Eight months later, two giant PC manufacturers are in dire straits -- Hewlett-Packard recently announced laying off over 27,000 employees and Dell's Q1 2012 earnings have been weak across the board in their Consumer, Public Sector and Enterprise divisions.

Apple, on the other hand, is doing magnificently, with their products accounting for over 22.5 percent of mobile PC shipments globally in the first quarter of 2012, 80 percent of that being their own Post-PC iPad tablet, with 17 million mobile PC units shipped, according to NPD Displaysearch.

The hard truth is facing us -- traditional PC purchases are slowing down dramatically. Unless you are cultivating a strong business in tablet computers and smartphones as well it's going to be a very uncomfortable ride in the next few years for the PC manufacturers.

And dare I say it, a number of them are going to go extinct.

So what exactly is Post-PC, anyway? Is it simply just a buzzword, or is it a legitimate phenomenon?

To put it bluntly, the Post-PC world represents a displacement of computing from the traditional, 30 year-old Intel architecture used on desktop to the Datacenter and the Cloud.

In essence, we are returning to a very similar highly centralized model that was popular in the late 60's and mid-1970's with mainframe-based computing.

The only difference is that instead of a monolithic, purely mainframe-based time sharing model, our new centralized architecture is multi-vendor and heterogeneous, can be distributed within Public and Private Cloud infrastructure in multiple datacenters and is more business resilient and more flexible than ever before.

Within ten years, the majority of business professionals will be using extremely inexpensive thin notebooks, tablets and thin clients (sub $500) which will utilize any number of software technologies that run within the browser or will use next-generation Web-based APIs and Web Services (Such as those available in Microsoft's WinRT and other HTML5 frameworks) to provide line-of-business application functionality.

In addition, I see the application programming standards used by today's most popular mobile operating systems -- iOS and Android -- being used heavily in business environments to provide the front ends to these Web APIs.

As soon as three to five years from now, the average business professional will be transitioning from "Heavy" clients such as desktop PCs and business laptops with large amounts of localized storage and localized applications using Intel chips and Windows to very small and extremely power efficient, SoC-based systems using completely solid state storage (SSD) which will function mostly as cache for applications that run remotely.

I see a mix of both ARM and Intel's next-generation Systems on a Chip (SoC) using sub-22 nanometer manufacturing processes fulfilling this role, with the predominance of the market being addressed by ARM-based devices as we move further into the future and backwards compatibility no longer becomes as pressing an issue.

I also expect that during the 5-year transition period to fully Web-based apps, there will be a significant amount of virtualized desktop apps (VDI) via Microsoft's Remote Desktop Services and RemoteFX in order to bridge the gap.

BYOD (Bring Your Own Device) will also be part of this equation. I see mobile hypervisors such as Intel's Wind River, Open Kernel Labs OKL4, Red Bend's vLogix and VMWare's own Horizon product being used in conjunction with mobile software management solutions to provide security and partitioning for the employee that chooses to bring their own devices to work, particularly if they are Android, Windows Phone or Windows RT-based solutions.

I also expect that for iOS, Apple will either partner with major enterprise software vendors for providing a similar type of security partitioning, or buy/roll their own solutions in order to make their mobile operating system a first class managed citizen in the enterprise.

[Next: How the extinction will play out]»

Now that we know what the technologies that will replace traditional desktop computing will look like, who is poised to lead the transformation of personal computing?

Apple obviously has an extremely large lead in Post-PC devices with its own iPad, and everyone else is currently at a disadvantage due to its market penetration and the maturity of its developer ecosystem. Anyone who joins this bandwagon obviously is in a very good position to penetrate the enterprise and the consumer space as a software or services vendor.

That being said, while Microsoft's current mobile offerings have had a lukewarm reception in the consumer space, the company still has tremendous potential for maintaining its lead in the enterprise, given significant advancements with the upcoming releases of the Windows 2012 Server operating system, the latest incarnations of Office, as well as Windows 8 and Windows RT.

And while Google and its handset OEMs are doing extremely well in the consumer smartphone space, I do not envision a pure consumerization route for Android tablets in the enterprise a la iOS in the immediate future, at least not until some initiative is taken by the company to write or provide incentives for 3rd-party developers to write enterprise-grade tablet apps, or until they provide a good management framework.

Perhaps at Google I/O we'll see some new developments in this area which I believe Android is lacking, particularly now that the Oracle litigation and the Motorola acquisition is behind them.

Mobile Hypervisors I mentioned earlier will certainly help, but Google needs more than that to provide a compelling reason for enterprises to use Android over iOS or Windows RT.

This is not to say Android tablets and other devices will not have an impact -- they will, particularly in vertical market applications, perhaps more-so than iOS or Windows RT, due to the ability for Android to be modified by the vendor. But vertical for the most part is niche, and not within the scope of widespread adoption.

Additionally, despite it being a good idea, Chrome OS/Chromebooks so far have had very little traction in the consumer or enterprise market at all, due to the high cost of the devices and their limited functionality.

However, with a possible fusion of the Chrome Browser for Android, inexpensive ARM SoCs and the developer/hardware ecosystem of the Android OS, the Chromebook and Chromebox concept might yet still have some legs.

We now come to the unpleasant part of the personal computing transformation -- the companies that will undoubtedly suffer as a result of the sweeping changes that will occur within the next decade.

Any company which has had revenues that are heavily dependent on PC manufacturing and sales had better start thinking about diversification and eliminating redundant products.

Obviously, the big targets are Hewlett-Packard and Dell. Dell may still weather the storm because it recently bought WYSE, indicating a future emphasis towards smart devices, VDI and thin clients.

Microsoft as a company will continue to survive by transitioning a large portion of its software business towards the enterprise and server-based/Cloud computing.

The consumer and business conversion to the Metro UI in Windows 8 will be a slow one while enterprises retain a significant amount of Win32-based desktop software infrastructure, as its server software and Cloud business will be strong and even expand.

Microsoft has always had an active role in the development of the Personal Computer between Intel and the PC vendors, and has been proactive in dealing with industry changes and trends.

However, there can be no denying that with Windows RT, the company is hedging its bets between Intel and ARM, which is obviously making its traditional partners nervous since a huge amount of their revenue stream has been Intel and Windows RT's success is not necessarily assured.

The last time Windows ran on multiple desktop architectures was in the early 1990's, when Windows NT was introduced on the PowerPC and MIPS alongside Intel. This is a big deal.

That being said, the continued health of its traditional partners such as HP and Dell are not necessarily guaranteed, as I have explained above.

Intel, on the other hand, may need to make some difficult choices in the years ahead. I think there will clearly be some demand for Ultrabooks and Windows 8, but as to how deep it will be compared to tablets and traditional laptops running Windows 7 I cannot say.

The prices for Ultrabooks certainly need to go below sub-$1000 or even less to make an impression, because your average consumer laptop goes for about $600.

I can say, however, that by not investing in ARM-based semiconductor technologies, Intel doesn't have much of a Plan B in place. While their Tri-Gate manufacturing process has enormous potential, x86 isn't necessarily the best and only target for it going forward.

As I touched on earlier, diversification into Post-PC products by these companies and eliminating redundant products/streamlining is going to crucial in order to survive the transition.

If we look at Apple as an example, they only have a few basic types of Macs and Macbooks -- the Macbook Air in two sizes, the Macbook Pro in three sizes, two sizes of iMac, two variants of Mac Mini, and the Mac Pro desktop.

The Mac Pro is likely going to be discontinued or some amped-up Ivy Bridge version of the Mini or the iMac will end up taking its place (iMac Pro?)

It also wouldn't surprise me at all to see Apple further streamline and eliminate the Macbook Pro/Air and just have a single Macbook line sometime in the future merging the technologies of both lines.

Apple has no problem doing more with less, and its profits reflect heavily on that philosophy.

By comparison, your average PC vendor has a lot more flavors of system they are trying to sell. For example, HP has at least 5 lines of consumer laptops with multiple sizes within each line, and at least 16 types of desktops. This is madness, and it's no wonder the company is experiencing problems.

There will always be a need for the traditional PC, but it will become more of a niche market than it is today. Ten years from now, less than 10 percent of the current global PC using population, if not as low as five percent will actually require them.

It could be even less depending on how big technologies like VDI take off. Give a developer or someone in the scientific/engineering field a lot of back end server power on a private cloud and a professional monitor attached to a thin client, along with technologies like Microsoft's RemoteFX for server-side GPU rendering of virtual desktops, and the need for those big desktops could entirely disappear.

The biggest hurdle the PC OEM industry will have to face is the inevitable and painful consolidation. This is a process which began 20 years ago, when there were literally at least a dozen Tier 1 manufacturers as well as scores of Tier 2 and white-box vendors. Many of those companies no longer exist today due to mergers, acquisitions, and companies simply going out of business.

The white box and system builder phenomenon is also in danger of disappearing entirely, much to my own remorse.

Lenovo is likely to usurp HP as top PC dog within the next two years because it is not as fat a company, it's a virtual subsidiary of the Chinese government with huge domestic demand for their products, and has done well diversifying into Android and Windows tablets in addition to smartphones.

In addition to Lenovo getting bigger or even consuming weaker brands (Fujitsu/Toshiba/Panasonic's PC divisions or possibly even Dell) we may very well see Microsoft and Intel have stronger relationships with what we might now call Tier 2 or Tier 3 Chinese vendors, such as Asus and Acer, which are not nearly as margin sensitive as the Big Three.

It is also inevitable that a bunch of these Chinese and also Korean firms will consolidate as margins become hair-thin.

It also would not surprise me to see Samsung's PC business improve due to their diversification and their leadership in consumer electronics and component manufacturing business as well.

One could compare this to what is regarded as the most modern theory by developed by leading paleontologists as to what really happened to the dinosaurs during the late Cretaceous period.

Before the big asteroid hit, dinosaurs were already experiencing a period of difficulty due to climate change.

While there was a single cataclysmic event that accelerated their decline, and a huge amount of the animals died out within a few years, it is now known that a much smaller population of non-avian dinosaurs continued to live for hundreds of thousands of years after the K-T event because they became highly specialized and adapted to a cooling planet.

And other species of dinosaurs evolved into what we now recognize as birds.

You don't have to be Carnac the Magnificent to see the writing on the wall. It's painfully obvious that the industry has too many PC manufacturers and they have too many redundant products to sell, with a falling demand for their wares.

There's your climate change. The accelerated acceptance of Post-PC devices like the iPad is the asteroid.

The majority of consumers desire and require inexpensive, mobile devices that do the basics (Web, Email, Social Networking) along with easy to use apps that leverage data stored in the Cloud.

Conversely, businesses and the enterprise want devices that are low cost, power-efficient, low maintenance and easily manageable, with the ability to leverage applications and data running in their datacenters.

Only Post-PC devices such as tablets, ARM-based systems and thin clients (such as Google's Chromebox) can achieve this.

Sure, some people will still require the horsepower and complexity of the PC. But this debate was never about the PC itself going away; it's whether or not PC companies as a whole can sustain a business model on revenue largely dependent on PCs.

And to that, the answer is a resounding no.

Topic: PCs


Jason Perlow, Sr. Technology Editor at ZDNet, is a technologist with over two decades of experience integrating large heterogeneous multi-vendor computing environments in Fortune 500 companies. Jason is currently a Partner Technology Strategist with Microsoft Corp. His expressed views do not necessarily represent those of his employer.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Huh?

    Maybe for some, but not all. Certainly not Dell. I also don't see "heavy" clients dying out. Personally, I will never EVER use a cloud only OS, or thin client that doesn't give me control of the system or where I store my data. I don't want my data online, unless I specifically put it there.
    I'm not online 100% of the day, nor do I want to be. Not with the way cell phone companies are jacking up the prices of their data plans just to make more money.

    I see heavy clients evolving to meet the needs of users like myself. Give me a nice 13 inch mid range laptop running Windows and I'm happy.
    The one and only, Cylon Centurion
    • Sometimes...

      you don't have a choice when the organization you work for goes through a hardware refresh and you're not involved in the decision making. You're speaking as a consumer... and I agree with you personally, but my organization is already looking at Dell thin clients. Whether I want to or not, I may be using a thin client at work within 3-5 years.
      • That's an Entirely Different Thing

        Thin clients at work are an entirely different thing, and are not in the least a new thing. Right now at the place where I work, we are using PCs. However, in the time that I have worked in the IT Department here we have used thin clients as well. Different configurations are switched between over time because of varying factors and considerations.

        However, none of that really means anything to the typical personal computer user or even means that much for "the cloud" in general because most companies serve their own thin clients. Until it becomes normal to trust an outside company with your entire IT infrastructure (something I can't see either management or IT being behind at my company any time soon), "the cloud" is just another buzzword for the same stuff that's been happening for decades.
      • Wang anyone?

        History is repeating itself. Client - Server to distributed architecture to Client - server. This is nothing new and guess what will follow. PC's. I totally agree with history and you that having everything in the cloud is a dumb idea. Everyone else will figure that out too.
      • Wang?


        Those of us who remember the Wang VS (with fondness) regard it as a machine that was a long way ahead of its time.

        The VS had programmable terminals fully controlled from the server and can be seen as a forerunner of the thin client approach.

        How many other machines had a full source code level debugger in 1979?

        Moving from the Wang VS to client-server resulted for us in a huge drop in productivity and reliability.

        Wang didn't die because of poor technology, but because of bad management.

        Thin-client is the way forward for businesses.
      • CFWhitman, Cylon made no distinction

        That's why he's wrong.

        But I wish the little fellow good luck finding a new job once his shill contract runs out.
    • I'm with you

      Clouds in my world are the sign of bad weather, rain, storms and power failures, I'm never gonna trust my data to something that reminds me of bad things.
      • Clouds make rain...

        and rain is is a good thing. Without rain you die of hunger...
      • Clouds make rain... Huh?

        Rain is stuff clumping together and falling down. Do you want your data to clump and fall?
      • Clouds occur

        When a huge body of hot air meets a large quantity of cold liquid.

        As for example when a marketing department encounters a refrigerator full of beer. The marketing people "interact" with the beer and become "creative" (I believe that is the currently popular euphemism).
    • The future of devices

      ... is the transformer phone and thin client computing.
      We are going to trend from having multiple devices each with their own CPUs and storage to having a single device (a phone) that does it all!
      What will happen is we have this powerful phone that we can dock into a tablet dock and transform it into a full screen tablet. When we get home or in the office, we can dock it into a lapdock to use it like a desktop machine with a mouse and keyboard. We get in our car, we dock the phone and the dumb touchscreen in our car becomes a carputer.
      There will be no compromise as our "personal computer" transforms into the ideal device we need for every situation.
      Further down the track, this concept will extend to thin client computing where we just use the phone as a VM server and we just use (multiple) thin clients to connect to the phone. You will then buy dumb tablets which just becomes a wireless terminal to your phone. The phone does all the connectivity, processing, and storage. The dumb tablet just does input and output. You will also have a dumb laptop terminal which does the same thing in laptop form. When you relax in front of your TV, guess what? Your TV becomes the thin client/dumb terminal for your phone as well. When you go to a net cafe, you log into your phone using their dumb devices and use it just like your own dumb devices at home or in the office. The market will just become dumb devices like tablets, laptops, desktops, TVs, even carputer screens which run as a virtual machine on your phone.
      The term Central Processing Unit (CPU) will take on a new level of meaning as you will truly just have one CPU doing it all. Your CPU virtually, dynamically, and multiple-ly scale itself to whatever devices you want to use wirelessly. Everything just becomes input and output for the phone.
      Having all this is great, but we need an OS that can handle all these transformations and scalability.
      iOS? no chance. Sorely lacking in scalability, functionality, and Apple is too tight to open it up. Iphones will become the Nokia phones of yesteryears. It will still work, but lack relevance and adaptability as mobile computing evolves.
      OS X? I don't see a rosy future for OS X as Apple continues to dumb down the OS and apply makeup to the UI. Dropping ZFS was a significant bad move for Apple's desktop OS. It is getting technically worse from Apple and doesn't look like turning around. They can fool the ignorant, but for how long?
      Android? half a chance. It is probably possible to do most of this on android already. However, Android presents a compromised situation in so many areas that it is not mature enough to do this properly. The main concern is that Android apps are designed with limited intent and limited UI - they are basically phone apps. If Google could scale the Android concept and get Android developers to think big and design their apps to scale devices, we could get there with Android.
      Linux? mature enough on desktops, but doesn't scale down to small devices well. It is the opposite situation of Android. It needs a huge shift in UI and the fragmentation pretty much makes a significant shift unlikely.
      Windows 8? this is the OS that has the best chance of doing it all. It ticks all the right boxes. It is the ONLY OS with a UI and backbone that could stand a chance of scaling from a watchphone to a TV. Microsoft has thought about the points I have made about future devices and it shows in Windows 8. Whether you like it or not, Microsoft is onto a future with Windows 8 whereas the rest will just be tweaking a tired OS and unscalable UI and will struggle to match Windows in the long run. Unless something unbelievable happens with any of the other OSes, I see nobody catching up with Microsoft in the long run. Microsoft is many moves ahead of the competition and while there will be resistance in the short term, Microsoft has done their homework and will be the winner in the long run whether you like it or not.
      • You are on crack!

        "We are going to trend from having multiple devices each with their own CPUs and storage to having a single device (a phone) that does it all!" That will become symptomatic of SOME people, but not the vast majority of the computing world.

        Server farms (think big industry all the way up to google) will be on stable server-based platforms. The folks who'll maintain all of that IT infrastructure certainly can monitor it via a phone, but for actually serious work? HA, they'll be using a computer (read PC in some cases, full server in others).

        Windoze8? Seriously?
      • Modular Computing

        You are missing the point.
        Server side computing won't be replaced by personal devices. They will be there for cloud storage and processing. The devices you use to access the servers is what I'm talking about.
        There are 3 basic elements to personal computing: storage, processing, and input/output (or HID - Human Interface Devices). The devices we have today combine all these 3 elements into one device. What will happen is the modularisation of these 3 elements.
        The Storage part can be done in the cloud and/or locally. Locally, we won't be carting around hard disks in client devices but use a centralised mass storage system similar to NAS devices. Basically it will be a wireless hard drive/SSD or whatever. The trend will be to make it portable so you can take your modular storage with you. When you need more storage, you upgrade it without touching ANY of your devices or paying for a device with more integrated storage. Redundancy or backups? easy. Think of it as your personal cloud storage.
        While cloud processing will grow and become significant in enterprise, local processing will still require a device with a powerful CPU. The processing part will be done by your phone. This is no longer your phone but a personal computer. The mobile phone has become powerful enough to be a personal computer for most purposes. You won't need to upgrade your tablets or laptops or other client form factors. You just upgrade your phone and its processing power, GPUs, mobile connectivity, and other features carry thru to your dumb tablets, laptops, TVs, carputer screen, etc which all just function as the HID element of computing. We no longer need to spread our money over multiple CPUs on several devices. We just buy one powerful phone and everything is channeled thru it. We no longer have to update apps on every device or sync data as all the apps and data is channeled thru the one personal computer - the phone.
        The desktop PC box as we know it today is just an integrated processor and storage. It will still have its need in special purpose computing. However, the future of personal computing is local modularisation irrespective of what happens in the cloud.
        Windows 8 is more ready for modular personal computing than any other OS.
      • Your vision is relevant, but...

        I certainly hope the future will be based on open technologies, and not vendor lock-in to Microsoft with all the ugliness that causes. If the Linux community manages to focus their efforts on writing lightweight interfaces for complex tasks, simplifying instead of complicating, the system will be mature enough for both phone and desktop, and able to implement the future you envision.
    • Agree with Cylon

      My wife is an interior designer at a big, wealthy university. They got to do filing all day yesterday because, for whatever reason, the network connection went down in their office. I guess they were living in the "post-PC era"... they had to do the work that gets put off for years because they always need to be using their computers!

      Businesses can't afford to lose access to their livelihood. Similar to how the majority of businesses are left in the dark literally and figuratively in a power outage, companies can't do much without their internet connection. Imagine if all the phone system went down at a hospital or if the network crashed.

      Ask yourself this... weren't CDs supposed to be dead a decade ago?
      • CDs not dead a decade ago

        Cds were not dead a decade ago, but they are becoming that way now. Life support might drag them on for a few more years; but as soon as BluRay makes cheaper blanks many will be using them for personal secondary backup. I use a raid setup but would like something non-magnetic. Maybe if large Flash drives become cheap, then BluyRay will also mostly die out.
      • I wouldn't count on BluRay if I were you

        CDs have been around for almost 30 years. I doubt BluRay will get that far.

        Could be you'll be stuck with another dodo on your hands.
      • Someone who knows something!

        At least one person understands reality. I see so many people say the land line phone is dead and we should get rid of telephone poles because everything has been replaced by the wireless world but virtually every cell phone tower gets all it's data and bandwidth via digital carrier coming via fiber optics and copper T-spans from those same telephone poles. Having spent years in the data and communications industry I have seen more complete network failures because of bad weather or a big traffic jam. I've also seen people lined up 10 deep waiting to get to use the pay phone at a 7-11 because the entire cell and data network has crashed due to a traffic overload. Wait until one tech uploads the wrong data IP info and the entire network starts cross connecting all voice calls and the data networks drop packets. I've seen it before and it will happen again. On top of that the network gets more and more exposed to hacker attacks but everyone will say "But that will never happen to me".
      • CDs

        [i]Ask yourself this... weren't CDs supposed to be dead a decade ago?[/i]

        Aren't they? I haven't bought a CD in the better part of a decade..
    • Perhaps the most important point....

      ....though only mentioned in passing, was the infrastructure- and bandwidth- dependent nature of the post-PC model. If you're an organization in control of those factors, the thin client/server model makes sense, although there's no guarantee that it will be the best solution. If you're not, you have no defense against unreasonable costs for data transmission without local power. It's simple supply and demand. Demand for data services will be inversely proportional to local power.
      Lester Young