The problem of the UMPC

One year after the release of the first UMPC, it appears that no one in the industry really wants to talk about these devices anymore. Intel for example, excluded the term from its presentations to analysts on Thursday completely and has begun talking about MIDs instead. This development is especially interesting, as the company was heavily promoting UMPCs in various consumer scenarios just two months ago at the Cebit tradeshow in Germany. For example, we learned that UMPCs should take on the role of detachable car computers. That is unlikely to happen now.

So, what changed?

Obviously, we are in the midst of strategy shift. The reason: The industry has a lot more data about who is actually buying UMPCs today. Pankaj Kedia, ecosystem manager in Intel's ultra mobile group told us that 60% of all UMPCs were actually purchased by businesses and the initially targeted mainstream market showed very little interest in this new product category.

Kedia cautioned that every new product category needs time to evolve and that even the notebook needed 12 years to hit sales of 1 million units per year. He explained that a product such as the UMPC will require about two to three years until a supporting ecosystem with supporting hardware and software as well as interest from customers will become significant. However, one could argue that the created expectations outpaced what the industry could deliver at the time and market research clearly missed what customers the UMPC could attract.

In fact, the discrepancy between product design, marketing and customer expectations may have been the ultimate reason to scrap the original idea of the mainstream UMPC altogether. While customers would expect a $500 UMPC that runs Windows Vista's AeroGlass interface just as fast as dual-core notebook, a permanent connection to the Internet, a GPS navigation device, a slim solid state disk and a battery time of eight or more hours, the hardware and software of 2007 just isn't there yet - and isn't likely to get there in the foreseeable time.

While the marketing departments forgot to ask what is realistic these days and what is not, there appears to be also an issue with UMPC product management. The idea of the UMPC is a fantastic playground for innovation, but it actually is a lack of innovation that we saw over the past fourteen months. There was not a single hardware or software idea that was unique to the UMPC and that could take advantage of this form factor. Instead we saw efforts to simply create a smaller notebook (or a more capable PDA, depending on your view) with no extra benefit but extra inconveniences for the user. What the UMPC will be

While only the industry knows for sure, the UMPC looks like it has hit dead end. To turn it around, there are two possible solutions: Leave it the way it is and sell in smaller volumes to the users who have been purchasing the first two generations of the products or correct the product design mistakes and offer a more tailored product to the mainstream market.

Intel says it will go both ways. In that sense, the UMPC isn't really dead. I will have to make a prediction here, but as of now the UMPC feels like the Tablet PC all over again. It isn't hard to imagine that the UMPC will end up in a market niche and replace the Tablet PC, which once was projected to revolutionize the notebook. Will the mass market care? No. It hasn't care until now and will not care then.

For the consumer, it gets more interesting. Intel has MIDs in mind, with a price tag of about $500, much less memory than a UMPC, flash memory instead of a hard drive and a very lean operating system. Intel is currently designing a new 45 nm processor, code-named "Silverthorne" for these devices (and future UMPCs). Silverthorne is said to be about as fast as a Pentium M four years ago and cost about as cost-efficient to manufacture as a 286 CPU.

As a result, Silverthorne will not be able to run Windows Vista, according to Kedia. In fact, the MIDs Intel is currently showcasing are running Ubuntu Linux and not Windows. Kedia said that initial MIDS will definitely be running some version of Linux, but the company is also talking to Microsoft to offer "some" version of Windows for the MID.

As an example for an early MID, Kedia pointed to Apple's iPhone which is rumored to integrate several Intel components.

Does it make sense?

We will get a mobile device that is positioned below the sub-notebook one way or the other. In a world that goes more and more mobile, and users who ask for permanent Internet connections and entertainment on the go, a MID makes a whole lot of sense.

However, it will be a challenge to differentiate the capabilities of the MID from smartphones that receive more and more processing power with each new generation. With the MID not being capable to run and display Windows Vista - an operating system most users will be used to when MIDs hit the market - it will be difficult to highlight the advantages of the MID and there may not be enough difference to a powerful and similarly priced cellphone.

We do have plenty of pocket sized entertainment devices on the market today - from multimedia cellphones to portable audio/video players. It is a matter of time until someone combines all wireless and multimedia features and the simple addition of an Intel processor certainly does not automatically create a new product category.

Can a MID be successful and will consumers bite this time? Only time will tell, but half-baked products such as Sony's Mylo were not convincing enough to stir interest from the mass market. As it was the case with the UMPC, a simple shrink of an existing product (such as the UMPC) won't cut it. The MID will also need an innovative interface and new ideas to generate interest.

Maybe someone should ask Apple for help?