Features

Why Tablet PCs aren't selling well - it's the chip set

Reports that the new Tablet PC is a flop have appeared. But the experiment in computer slates is not really started yet, and the real test will come when the Centrino chip set is used. And, of course, when all the other known bugs are removed.

The "other bugs" aren't just price. It is still the case that you have to pay a bit of a premium to get a tablet, compared with what you'd pay for the same spec notebook PC - but the difference is smaller than it was. But what do you get for the extra money?

Increasingly, buyers are saying that you don't get anything extra; rather, you get something that is illegal, clumsy, out of date, almost impossible to use, non-standard, and often, badly flawed.

The biggest problem is that the world has really fallen in love with Intel's low-power PC design, the Centrino.

The announcement by Fujitsu-Siemens that all its notebook PCs would be Centrino based - made at last week's WLAN Event in London - had an important exception. That was the Stylistic Tablet PC.

The surprising fact was that the Stylistic tablet is the first to work properly out of doors. It has a transflective screen, which means it actually uses ambient light to provide the backlighting for the display.

So where an ordinary TFT display looks altogether dim and vague in sunlight, the Stylistic ST4000 actually works better in the open air.

But even so, the Stylistic is remarkable for being the only Fujitsu Siemens portable computer now still using a Mobile Pentium III processor chip. Why? The Mobile PIII is a known power hog, and was produced for those who found the standard Celeron too slow and were prepared to put up with poor battery life in exchange for a faster notebook, about two years ago.

"The problem is that corporate users won't commit to a new design without testing it," said Andy Barker, Fujitsu Siemens mobility boss. "They have what they call a "gold image" - a template for how they set up corporate PC hard disks, a disk image that these people already have on other notebooks."

Barker observed that unless they can just install their current gold image on a new machine, they can't test the device simply.

"These machines are being considered for mission critical applications, data collection in the field," said Barker. "They can't be debugging a new machine at the same time as getting their job done."

So, from his point of view, it was more important to get a machine out which actually worked - both on existing disk images, and also out of doors - rather than to have one which provided some other marginal advantage. Once the evaluation phase is over, a Centrino based Tablet can be brought to market, he said.

Are initial sales all evaluation machines?

"We wouldn't call them market research, no," disagreed Aaron Fright at Viewsonic. "But it is true that most of our Tablets have been bought in order to see if the slate format works. And it's definitely the case that we're getting good feedback from these customers about what else they want to see in a Tablet design."

One particularly hard lesson Viewsonic had to learn, was that the touch sensitive screen of its previous slate designs, is very different from the new electro-sensitive display on the Tablet. The wireless antenna, which works fine inside a slate like the Smart Display (touch screen) seems to be screened from radio; and the wireless network performance is poor. Users have actually had to disable the internal 802.11g WiFi circuit, and fit a PCMCIA standard card with an external antenna.

Demand for an expanded battery with extra capacity has also been higher than predicted.

There are so many variables, that it really is hard to know what the magic formula is. One thing, though, is clear; and that is that Microsoft and its partners have based the design on American criteria.

The biggest obstacle to European sales isn't poor visibility in daylight, or miserable battery life, or even dodgy wireless performance. It isn't even price, though that hurdle has yet to be cleared, and may prove the killer. The problem is a keyboard, and European regulations on usability.

For street use, no data collector wants the extra weight of a qwerty device, and machines which - like the Acer Travelmate - which turn into slates, but keep the keyboard attached, are seen as a compromise. They want a pure slate.

But nobody wants a pure slate when they're back at the office; they want a PC. Here the alternative is to provide a stand for the slate, and attach a keyboard and mouse via USB. It then looks just like a desktop: Viewsonic offers US customers a free mini-keyboard with the V1100 tablet; the stand illustrated is extra.

Like most attempts to turn a tablet into a desktop, it falls foul of European regulations on tilting and swivelling displays for office use; no corporate customer could consider buying these as they stand. Another reason, of course, to experiment on a small scale, and get the design right, before going into volume production.

Even then, there may be problems.

Turn a flat panel display through 90 degrees, and you completely change its visibility characteristics. Normally, TFTs are designed to be viewed in "landscape" mode, with the assumption that everybody will be sitting with their heads at roughly the same level, but that they may cover a pretty wide range of angles. It's taken for granted by notebook users, who have learned to make sure the screen is tilted so that it is pointing straight at them.

But stand the thing on its side, and suddenly, only the person sitting bang in front of the screen can read it. Admittedly, they can have good screen readability even if they stand up, or crouch down - but people tend not to do this much at desks.

The biggest change will be the adoption of Centrino technology. Acer has already switched to the Centrino platform, with its TravelMate C110 at $2,200 described as "the world's lightest convertible" Tablet. Fujitsu Siemens will follow, probably later this year.

Nick Eaton, marketing for Fujitsu Siemens pointed out that the Stylistic isn't their first venture into tablets. "We have a lot of installed users on pen tablet technology, and it's really not new for us; you could call this our 20th generation - we've done this for 12 years."

Eaton says that he does want to move to all Centrino - and it won't be just the main chip set, or the processor, but it will also be the Intel Pro Wireless circuits. Admittedly, the next tablet, A5 in size, will be based on old Windows CE technology - the same circuitry that drives most Pocket PCs, really - but after that, there will be a Centrino based one.

Other vendors have tried to compromise in other areas. Hewlett-Packard, for example, decided to go with low power as the essential ingredient, using a Crusoe chip instead of an Intel one, rather than wait for Centrino to ship. The "Compaq" design is one of the most popular as a result, but the processor hasn't got the "grunt" of rival machines. One sign of this is the decision to disable one of the Tablet's nicest features - the ability to sense how hard you're pressing the stylus.

Hardly surprising, you might think, with all these different drawbacks and question marks, that Tablet sales are sluggish.

It does seem to have come as a shock, though, to some observers. DigiTimes reported that growth in Tablet sales has been minimal, with Acer and Hewlett-Packard seeing the number actually decline.

IDC research appeared to show that HP sold 17,000 tablets in the last quarter of 2002, beating Acer with 15,000. But "sources" quoted by DigiTimes say: "HP had prepared component inventories to produce 30,000 Tablet PCs in the fourth quarter. Since then, HP has ordered a smaller amount of components for its Tablet PCs than it originally planned, according to HP suppliers in Taiwan."

Sales are definitely down. But it's wrong to suggest that the format is unpopular.

Users who have switched to the Tablet are often wildly enthusiastic. Stories abound of employees who have been given a Tablet to test, for a short term period, who have simply refused point blank to return them to the IT department at the end of the period.

This could become more marked, soon. Microsoft has an application called One Note, which will be shipped as part of Office 2003, later this year. Office is currently in its second beta-test phase, and Beta 3 isn't expected for another couple of months, so One Note won't hit the streets till September or October as a best guess; and it could easily be later.

One Note transforms the Tablet from "a notebook without a keyboard" into "this is how I want to work" for many people. It not only allows hand-written notes, but allows them to be made the way people tend to think - all over the place.

"Where Journal on the Tablet turns your notes into a serial stream, One Note allows you to cluster items all over the page, without putting them in an artificial queue of ascii," said one reviewer.

It also allows you to record meetings, and annotate the audio. The audio recording feature will run out of battery long before it runs out of storage, and every time the display is touched - even if it's no more than a dot or a scribble - that mark is linked to the audio track. You can write down little more than who was speaking, and what the subject was - or highlight items of interest - then later, play back only those parts which you highlighted.

Observers do think that this will create a market for the Tablet iself, rather than just allowing it to sell to a sub-set of the notebook market. Other applications will, no doubt, appear from other sources - specialist market research check applications, for example.

Microsoft itself hasn't yet seen fit to make the Tablet a standard. When it replaces Windows XP with Windows Tablet Edition, including all the extra features in its normal desktop offering, it will be clear that it is utterly committed to the Tablet concept. But the fact that it isn't yet doing this probably doesn't mean it is holding back.

Like everybody else in this market, Microsoft wants to get the market research phase right, and make sure that if there are extra features it hasn't thought of, they are supported at the deepest level of the operating system and application suite. That will take time.

In the next six months, definitely, at least two Tablet manufacturers will quietly drop the concept, and others will probably stop trying to manufacture their own designs - and will switch to a "generic" standard Taiwanese design which they can customise, as they do with desktops.

But don't assume that these are signs of a collapse in the market, It's almost certain that this time next year, there will be a spurt in enthusiasm for the Tablet concept, and a growth in sales - as the Centrino based designs start making it possible to have "all-day" battery life for workers in the field.

"We're already seeing huge commitments to orders," said Nick Eaton. "We can't reveal names yet, apart from Pepsi, which uses our Tablet for all mobile sales people, and has redesigned its workforce around the machine." But he expects more orders like that, as mobility and wireless start to become part of accepted thinking.