The Future of the PC
p2pnet.net News Feature:- The world spends an awful lot of time wondering what software will be in future PCs, but not so much time figuring out what they will look like, what they might cost and different ways in which the enterprise might decide to put them to work.
There has been a tendency over the last three or four generations of the Windows operating system to never consider anything, at least at the client level, other than upgrading to the prevailing version, regardless of what hardware that decision necessitates.
Enterprises now are just on the verge of starting to think about using Linux for handfuls of desktop activities, and in some instances converting off the Windows beaten track altogether and jumping onto the Linux bandwagon whole heartedly.
And by doing that Microsoft will lose some of its ability to decide just what a PC will need in hardware terms to sell in the market as an enterprise device.
But lone thing is for sure, that bulky desktops will over time become a thing of the past. Desktops have only survived this long because various Microsoft options have required big upgrades in computing power, memory and disk, and manufacturers couldn’t fit them all into a small enough box to give us all laptops with the same kind of power for a small enough price.
That is now changing and the laptop is clearly, rapidly on the rise
In 2007 notebooks will make up around 65m of the 175m PC devices shipped. Not quite a majority, but well on the way to 40% of the market.
One of the commonest terms in laptops now is “desktop replacement” which some companies abbreviate to “DeskNotes.”
At the moment these devices weight eight pounds or more and have 15-inch to 17-nch screens, composite keyboards, at least 80Gb of disk, plentiful USBs and usually rely on a docking station for much of their higher speed communications.
They have finally got the price substantially under $2,000 and now include dual CD-DVD combo drives, both writers and re-writers, in-built (Centrino) wi-fi and more disk than you could use on email in a lifetime.
But increasingly there are laptops that you really can consider carrying about with you, both inside an office and home and back every night.
These come in gradations all the way down to under 4 pound devices which sacrifice a little on the screen size and end up at about 10.5 inches, they have a little less Disk, which is usually a bit slower, along with a slower processor, in order to save power and extend battery life. But is that the end of where PCs are going?
We should be able to bring that size device up to desktop PC specification levels in just a few years, but is that it?
Well let’s dig underneath the motivation for laptops first, to see what clues this might give us.
What component suppliers are doing to drive the form factor downwards is usually paid for out of some kind of measurable productivity improvements.
These come in a number of ways; allowing work on the move; allowing connections on the move; allowing employees to work at home and allowing an “always on” email (or other “presence”) connection .
Rethink spoke to the CIO of Intel in a previous article and he told us how he had been totally against mobility for years, but ended up insisting on it.
The proving point for laptops was when he discovered that the cost of ownership was $800 more per year for laptops, and that this amounted to 30 minutes extra work a week per employee, in order to pay it back. When he measured laptop worker’s outputs, he found they “donated” between 3 hours and 8 hours of their “free” time to the company (even if most of it was commuting time to and from work) when they had laptops.
Taking this into account enterprise wide, it would mean that more could be done with less employees when they all had laptops so he gave them to everyone.
There were similar cost calculations when PCs came in initially, and the sight of a lonesome office worker, queuing up for his or her “go” on the PC is now a picture from way back in our past. It seems silly now, when we look back on it, to employ someone who is, in many cases, essentially an information worker, and then deprive them of the information they need to do their jobs.
In large enterprises that’s almost everyone, and the corollary is also true, that the greater part of their time that they have ALL of their corporate information available to them, the MORE productivity is obtained from enterprise employees.
This is perhaps why there have been more and more experiments with the PC form factor. In 2002 Microsoft pioneered the tablet PC. This was pretty much an unmitigated disaster, and remains less than 1% of PCs shipped today.
The same rules applied to these as applied to laptops before them. The smaller something is, the more expensive it is. And within that form factor the faster devices which perform closer to the desktop, cost even more.
So for two years now people have “bleated” about the tablet PC being too expensive, and that it doesn’t have enough applications and that when this is fixed, it will come into its own.
The difference here was that tablet PCs were not only normal PCs that were smaller, but they were also re-designed to work differently. Once a device gets too small there is a temptation to do away with the keyboard and use a stylus for input and once you do that there’s a temptation to use handwriting and both store it and try to interpret it using recognition software. It’s then that you realize you have a brand new architecture and you don’t have the applications for it.
Only this month Sony introduced its U Type Vaio into the US, a fully fledged XP device which weighs only a pound and which looks more like a handheld games device (see picture below)
It features a 5-inch SVGA display, with touch screen and comes bundled with its Sonic Stage online music store software so that whoever buys it, consumer or business, they can listen to music when they are bored.
It allows the screen to rotate and to zoom in and out.
Apart from that it’s a typical notebook PC with a 1.1GHz Intel Pentium M 733 processor, 512MB of RAM, a 20GB, 1.8-inch hard drive, wi-fi card. It’s just that instead of weighing 4 pounds with a 10.5 inch screen, it is about a quarter of the weight with a much smaller, higher quality, screen.
It has a fold up keyboard and a tablet style stylus for handwritten notes, but does not run Microsoft’s tablet software and starts at around $2,000 to $2,500.
One reviewer joked that when he carried it around his house he almost forgot he had it and that it almost fell into a full bath, which would have made the subsequent data loss a lot more expensive than the price of the device alone.
So the PC vendors are playing this guessing game of trying to add new function as well as soldier away at making the device smaller and smaller.
Laptops could have a scanner added perhaps, a miniature LCD video projector, a camera for video conferencing and stills. These could all be added to that Vaio in about 18 months and still not add to its weight.
What is it that enterprises are after? Is it a PC that has a full sized keyboard, and a 21 inch screen, and yet folds up to fit in your pocket, holds enough disk to keep a library of High Definition TV films and which has broadband communication from anywhere. Oh yes, and perhaps it should costs about $500.
Well actually that would be very handy, even if it is a contradiction in terms, having to be both large enough for comfort work and small enough to fit into your pocket. Because that’s the only way that all of the productivity benefits are achievable, consistent access to corporate information, easy to work on and small enough to take into meetings, unobtrusively and light enough to carry around all over the world.
At the moment we are faced with two devices, if not three, to give us all of those benefits. We need a PDA so that we have a small form factor that we can take into meetings or carry in a pocket, we need a mobile phone to keep us in constant communication with people that want decisions out of us and we need either a laptop or a desktop (or both) to give us a convenient method of working when we are in between home and work, or traveling or when we finally get home.
But the costs and the cost of ownership of three devices is naturally prohibitive in the enterprise, not to mention defeating the object. When people have to sync two devices or move email between two devices, it has its own drawbacks, so the technology world is trying to find the elusive form factor which will meld them all into one.
A number of developments coming down the track point to how this impasse might be resolved. While we certainly have disk drive form factors getting smaller rapidly, and screen manufacturing processes and higher resolution makes it easier to live with a smaller screen size, it is communications technologies where we might find the answer.
The Multiband OFM grouping led by Intel and Texas Instruments and the competing faction driven by Motorola, are both working towards a suitable offering that will provide the foundation for both wireless USB and Firewireless.
A short range Ultra Wide Band signal offering 3 meter hops with data rates of up to 480 megabits per second, is initially really only supposed to give us a way of doing away with cable in the office.
Wireless USB will certainly be a way of connecting things to a PC without wire, including the keyboard, the mouse, any corporate network and in turn central storage facilities.
This concept is heavily influenced by Intel and Intel is backing Multiband OFDM to carry this protocol.
Chips go volume in mid-2005 and they will only put $50 on the price of a PC. Intel also hopes to build them into Consumer Electronic devices for the home, which means this has the potential to become a 1bn device per year market. This is turn will take the chip price down below $10 rapidly if it works out.
PC peripherals will present themselves rather like a wi-fi connection does now. There will be software selection and auto selection of devices. The printer will wake up WHEN you push the Print button. The communications port will authenticate WHEN you load Internet Explorer. There would be no further need of docking stations.
If you think about the various elements of a PC, the disk drives, the screen, the keyboard, the internal bus, the mouse, the speakers etc… and imagine them for a moment mostly broken into parts and talking over a 3 meter distance, suddenly you have a very different view of what the PC actually is.
Instead of its being a set of interlinked components, it becomes a set of capabilities. The capability to save a file, to open a file, to amend a file, to listen to something, to watch something on a screen, to input something.
Depending upon which devices are present, those are the capabilities that you have available to you.
But those devices no longer have to be wired quite so tightly together and can, if necessary, be broken up for use separately.
It’s not that this has been overtly placed on a roadmap by any of the PC manufacturers, or that Microsoft has intimated that it will adapt its operating software to allow this, it’s just that this solves all of those issues about building a machine that we have seen contradicts itself.
This Wireless USB concept has been driving another, related concept which has taken hold at Intel, which it calls the Personal Data Server. This has two forebears in the small, large capacity Flash memory drives that are currently being offered as a way of taking work home. Soon we will be able to copy up to a Gigabyte of work on Flash and carry it home and then bring back in the morning. The other forebear is the iPod, which has demonstrated to everyone that Micro drives, which were first applied to much the same application, are now viable and getting cheaper every day, and these carry so much more than Flash memory.
The issues are whether or not users have the operating environment at home to continue working. One or two companies are even lobbying Microsoft to get permission to take the operating software with it and store it on the same piece of Flash memory or removable disk drive.
But these key ring sized devices sit inert in our pockets from when we leave the office until we get home and they have no constant communication facilities or a small interface for work on the train or offering “presence” in a meeting.
But if we add a long distance communication service like cellular, or a broadband service like WiMAX to the wireless USB, this concept improves dramatically.
Suddenly it’s simply that the hard disk drive Microdrive forms the heart of the architecture and is detachable (in fact it never has to be attached, just within 3 meters) from your main PC. The drive itself behaves like a Research in Motion device, with a screen and stylus input, offering email and instant messaging and any other form of presence you want (paging if you like). In this format it could still be used for “proper” work, if you can put up with the tiny screen for a while, so good for emergencies.
Because the device will need a WiMAX chip in it, it has a broadband connection and with a simple earphone headset can be a mobile VoIP station for telephony, it is either a mobile (after 2007) or at least portable, phone (operator permitting).
Intel has promised to put WiMAX into its mother boards from next year, and the mobile version as soon as it’s available in late 2006.
For those that have not heard of WiMAX, it offers a shared 75 Megabit per second stream of broadband over cellular distances, with much quicker and cheaper network rollout than cellular.
Some devices, we are sure, will also include portable Wireless USB keyboards for work on the train, or in hotels, and perhaps even Wireless USB screens. It might be that the Portable Data Server as such sits within some form of cheap laptop device for working in a portable environment. Naturally these would initially have wi-fi too for all the benefits that this gives.
This tiny 1.8 inch or 2.5 inch hard disk becomes almost like an aircraft’s little black box flight recorder for the PC. It’s heart and sole is wherever that lives, along with all of your corporate data.
In a customer’s office the hard disk simply requires permission to authenticate with any other screen and keyboard and it becomes a desktop. In a hotel business centre much the same, and at home the same again.
On the road a small touch screen and stylus on the Portable Data Server turns it into a mobile email/ present device with a phone attached.
For liberal enterprise environments a segment of the drive (or yet another very tiny drive) operates as a Personal Entertainment Server and when the office worker has had enough of work on the seven hour flight to Europe from the US, he leans back and uses it as an iPod.
Disk drive market leader Seagate has shown the way here, recently announcing that it was moving away from the price sensitive 3.5 inch form factors towards 2.5 inches for laptops and down to 1 inch for consumer electronics devices.
It thinks it can make good margins here and yet it has already sparked a price war in the 1 inch space, where it offers a 5 GB device. This will be at the 20 GB level in three years, with virtually no price increase and then we will begin to see the prospect of a Portable Data Server in a mobile phone sized form factor.
But in perhaps a year, the first portable data servers will emerge and later will take on many shapes, from iPod style stylus devices to laptop weighing a couple of pounds, but the underlying principles will still be the same.
And as for the last part of the requirement, that it costs just a few hundred dollars, so that when you lose it, drop it in the bath, or leave it on the train, it can be easily replaced, why should it cost much more than $500, after all its mostly a disk drive.
This won’t reduce total spend on these devices, because of all the extra keyboards and screen frameworks needed, the multiple wireless USB chips and additional services like new security and fractional remote backup system which run over WiMAX, so that when you do flush your Portable Data Server down the toilet, all that really valuable corporate data hasn’t really gone with it.
Peter White – Rethink IT, UK
December 5th, 2004 at 4:23 am
Where is the graph you mentioned? You wrote, “But looking at the graph below one thing is for sure, that bulky desktops will over time become a thing of the past. ” I don’t see a graph.
Also, a Tablet PC running Microsoft Windows XP Tablet PC Edition / 2005 can run all Windows XP Home/ Professional applications plus applications that take advantage of digital ink and advanced speech recognition’. So, when you compare it with the Sony, a Tablet PC can actually do more and you can get one for less money. (Check out Acer C112, for instance, it’s less than $1500 US.)
December 5th, 2004 at 8:50 am
Peter,
Very good, well-researched and thought-provoking article. But I have to take issue with your view of the Tablet PC. It’s not that your view skews the article, not at all (it’s insignificant relative to your overall message). My objection is that you’re spreading misinformation about the Tablet PC.
From Microsoft’s point of view, the Tablet PC is NOT an “unmitigated disaster”. They were quite happy with the ~400K units shipped in 2003. What’s missing is an understanding of WHY Microsoft created the Tablet PC in the first place. It was NOT intended to be a “totally new mobile platform”, regardless of what their PR and ad agencies said. (Those message were just to get the product started. Already Microsoft’s messaging is starting to move away from the theme of extreme portability towards mainstream usage of the pen.)
Microsoft has a simple but hard-to-achieve goal: add a third input device to the PC. Today the two primary input devices are the keyboard and the mouse/touchpad/pointing stick. But these devices aren’t very good at some things, such as (a) sketching, (b) annotating documents, and (c) taking digital notes without making noise. These three things highlight the value of the pen. Without a pen, you can’t do these things very well, especially while mobile.
So, how should Microsoft go about add another input device to the PC? Just tell the OEMs to add it like Firewire? It’s not that simple. The pen needs software support in the OS and from ISVs. Firewire and many other hardware additions don’t. This means that Microsoft can’t just cram the pen down the OEMs’ throats like they normally do with new hardware.
As a solution to the problem, Microsoft created a mobile platform where the pen makes sense (the Tablet PC). The role of the Tablet PC is (a) get the world used to the idea of a pen on a PC, and (b) start ISVs modifying and developing software that takes advantage of the pen. An example of the latter (that hasn’t happened yet) is Adobe Acrobat, often used for annotation of documents. Acrobat doesn’t currently support Microsoft’s digital ink standard, but it will eventually.
Microsoft doesn’t sell PC hardware. They don’t have any allegiance to any particular hardware form-factor. They don’t care how many OEMs get burned creating new hardware platforms that never go anywhere. (Just think about the three generations of Windows CE hardware that were born and died before the iPAQ hit the sweet spot!) To Microsoft, the Tablet PC is just a means to an end. If the sales stay in the “few percent” range over the next few years, that’s perfectly OK with them. They are gradually accomplishing their objective. They take a MUCH longer view of this than anyone ever realizes, especially the media! Writers like you and many others believe that if a new product doesn’t take the market by storm and rack up more than 10% penetration in a year or two, it’s a failure. With a consumer product such as the iPod, that’s probably true. But the Tablet PC isn’t an iPod. It’s a different beast with a different purpose in life and a different lifespan.
The next piece of the puzzle is Longhorn. Today, to write a pen-enabled application you have to do special stuff — you have to use the Tablet PC SDK. The pen isn’t integrated into the OS, it’s really just an add-on to XP Pro. Longhorn changes that situation drastically. Take a look at an architecture diagram of Avalon (Longhorn’s presentation subsystem) and you’ll see that the pen (digital ink) is fully integrated everywhere. When Longhorn is released at the end of 2006, there will no longer be a separate Tablet PC OS because it won’t be needed. Microsoft believes that the four years of Tablet PC hardware development (2002-2006) and the steadily increasing number of pen-enabled applications from ISVs will pre-dispose the market towards accepting the pen in Longhorn. If the pen is totally integrated in Longhorn, and if OEMs know how to build a pen into a laptop (which they will have been doing for four years), why wouldn’t both ISVs and OEMs start to consider how & where their products could take advantage of digital ink?
By the end of 2009 (three years after the Longhorn launch), Microsoft expects that the percentage of standard laptops that incorporate a pen in some form will be somewhere between 25% and 50%. There won’t be any “convertible Tablet PCs” in 2009 because a laptop with a pen will be mainstream. That’s not to say that all laptops will become convertibles, not at all. Consider that if all you need to do is sketch and annotate, why not allow your laptop screen to open 180 degrees and lie flat on the table? If the laptop has a wide-angle screen (160 degrees all around), using the pen is perfectly OK with the laptop flat on the desk. That’s one of the reasons that Microsoft is pushing the LCD vendors and OEMs so hard towards wide-angle screens.
End of 2002 to end of 2009 — that’s seven years! Do you see why I say that Microsoft is taking a very long-term view of the Tablet PC? And why I say that the Tablet PC is NOT a new form-factor, but instead simply a vehicle to help drive the pen into all PCs?
Why does Microsoft care about adding the pen as a third input device in all PCs? Listen to some of Steve Ballmer’s speeches and you’ll hear him say, “We want you to use the PC 16 hours a day!” He doesn’t mean sit in front of a screen for 16 hours, he’s talking about integrating the PC (in many different form-factors) into every aspect of life. In the office, in the car, in the living room, in the recreation room, on the road, on your wrist, in your pocket, everywhere. What’s missing that’s needed to enable that integration? Actually, many things, including OS reliability, simplicity of operation, security, etc. — but that’s a different story. What’s also missing is “more input devices”. You simply can’t do everything with a keyboard and a mouse if you’re going to integrate the PC into your life. Pen and speech are two critical input devices that Microsoft has been working on for many years. Speech is coming along, but it will still be 5-10 years before it meets the average user’s expectations. Pen is much closer to being ready. There are some issues, like the cost of the required active digitizer, but Microsoft is working hard at getting that cost down (for example, by pushing other digitizer vendors into the market, to break up Wacom’s monopoly).
That brings up the subject of handwriting recognition. Three paragraphs after your “unmitigated disaster” comment, you talk about “doing away with the keyboard and using the stylus for input”. That is NOT Microsoft’s vision or goal. The pen IS NOT a replacement for the keyboard, and it will never be. Handwriting recognition is steadily improving, but it’s still nowhere near accurate enough, fast enough and easy enough to correct. Nobody will consistently use the pen instead of a keyboard, except in two special situations. The first is taking notes in a meeting when it’s rude to use a keyboard. The reason for using a computer rather than paper in this situation is simple: take the notes in digital form so they can be searched later. It’s not necessary to explicitly convert them to text — after all, how often do you transcribe notes from your spiral notebook into Word? Probably less than 10% of the time. As long as the notes are searchable (via the “magic” of background recognition, of which the user is never aware), that’s all that’s needed. Wouldn’t you love to be able to search the notes in your last ten years’ worth of spiral notebooks?
The second situation is when you’re using the computer on the couch, or reading in bed, or whatever, and you need to enter a file name to save some notes, enter a URL, or scribble a really short handwritten email, etc. This activity uses explicit handwriting recognition, and it must work very well for users to accept it. But it’s NOT replacing the keyboard. It’s just a convenience for when the keyboard isn’t readily available or convenient to use.
Look at Microsoft’s ads on the Tablet PC. You won’t see much mention of handwriting recognition. You’ll see lots of mention of digital ink, which is what the pen is really all about. Sketching, annotation and note-taking are the three key applications for the pen. I’m talking about horizontal users (knowledge workers and consumers) here, not today’s vertical users such as nurses and insurance salespeople.
One final point. So far, the Tablet PC has been the only form-factor in which Microsoft would support a pen. But they recently and quietly announced that OEMs could now sell the Tablet PC OS on a desktop, as long as the desktop package included a monitor with an active digitizer (a so-called “tablet monitor”). If you look at the big picture, this makes sense. It’s another step on Microsoft’s path towards integrating the pen into all PCs.
In summary, the Tablet PC looks like “an experiment with the PC form-factor”, but it’s really not. Its purpose is as a vehicle for introducing the pen to the PC ecosystem. It’s one step on a 7+ year journey towards adding the pen as a third input device on all PCs.
Geoff Walker
Walker Mobile, LLC
Email: geoff.walker@att.net
Web: http://www.walkermobile.com
December 5th, 2004 at 9:38 pm
On the graph – sorry: that shouldn’t have been there.
Cheers!
January 7th, 2005 at 12:44 am
Wireless USB . USB Memory . P2P file sharing and media services. Your analysis is comprehensive and astute.
Memsen offers a device scheduled for production in 2006 that will compliment your future of the PC scenario quite nicely. See the Click n’ Share Keychain at http://www.memsen.com for a look at true mobile file sharing capability and its impact on real life digital file sharing in mobile settings. We will now move from PC to PC file sharing only to true portable device to device sharing with Memsen’s Ultrawideband products.