Technology: December 2008 Archives

Apple uses some Imagination

19 December 2008

It's tough to disagree with Rik Myslewski at The Register's take on Apple's decision to buy around eight million shares in Imagination Technologies. Although Apple is already using Imagination's PowerVR through the Samsung applications processor in the iPhone-class devices, all the signs point to Apple wanting to shift more of the software burden to graphics processors in not just its desktop products but the mobile ones as well. This is why the Snow Leopard release of Mac OS X is likely to be more important in the future than the laundry list of features that characterised its predecessor, Leopard, as I wrote in The Guardian yesterday.

For anything that can make use of large quantities of parallelism, graphics processors (GPUs) are more energy efficient than general-purpose processors. In fact, they are so good that we may well see a subtle rebranding of these devices, underlining the idea that do not have to be used just for graphics. The term "stream processor" is turning up more often. Take nVidia, which makes the GPUs used in the latest generation of MacBook machines. The company refers to the processors inside the chips as stream processors, a nod to Prof Bill Dally of Stanford University who has come up with a number of key concepts for doing efficient parallel processing and is working on some ideas for improving their energy efficiency.

Mobile strategy: SNAFU

10 December 2008

For the past two-and-a-half days, my mobile phone has been in clear sight of a basestation but for most of that time, been unable to connect to the network. There's nothing wrong with the phone. It can find the Cingular basestation at Isla Vista, just outside Santa Barbara, California. The basetation won't let it log on. That's because I have a T-Mobile contract. Move to the top of the hill, and the phone finds the local T-Mobile basestation.

It turns out that the two companies have a roaming agreement but it doesn't operate where the companies' coverage overlaps. Welcome to the bizarre world of mobile comms where a combination of greed and stupidity means these companies would rather lose money than possibly contribute to their competitor's top line.

You can explain a lot about mobile comms using examples like this. For years, they have operated a beggar-my-neighbour attitude that is only overcome through absolute necessity or regulation. In the early part of this decade, we could guarantee stories about mobile stupidity more or less every week as they tried to push multimedia messaging (MMS) as the supposed high-profit alternative to SMS.

Engelbart's big knees-up

9 December 2008

As mouse companies all around the world celebrate the intellectual laziness of the computer industry better known as the 40th anniversary of Douglas Engelbart's demonstration of the mouse, we should be thankful that the pioneer of graphical user interfaces rejected an earlier idea for moving a cursor around the screen.

So that users could keep both hands free for the keyboard, the original plan was to put a lever under the desk that you would then move around with your leg. Can you imagine what you'd have to do to laptops to make that work?


The anniversary is also a reminder of the crushing tendency of the technology business to assume that one size fits all. It's a real shame that it's almost impossible to get a chorded keyboard similar to the one used by Engelbart alongside the mouse in the original demonstration. Companies like Microwriter tried bravely to build a business around them and the chording idea had a brief resurgence in the research community thanks to Thad Starner's work on the Twiddler at MIT and then at Georgia Tech, which you used to be able to buy from spinoff Handykey.

The problem is that the way the industry is set up is that it costs too darn much to build effective prototypes and low-run user interface products. It's a problem that plagues researchers who want to work on ambient intelligence and wearable computers as well. But there is, apparently, no technique in sight that will make it happen.

This isn't going to be a popular sentiment but if the Nesta £1bn early-stage fund comes off, it is almost certainly too much money. The problem doesn't like in cash so much as knowledge. If you create a fund with too much money, it runs the risk of flooding the market with well-funded companies that cannot survive long-term because their ideas are good but cannot sustain a business.

The trouble with early-stage investment is that it is clearly A Good Thing. So, more money for seed investments is clearly A Good Thing. But is it the right thing to do?

Venture capital in the technology sector has, for the past ten years, been a bust. At the beginning of the year, venture capitalist Andy Rappaport told a bunch of TSMC's customers - most of whom had been venture funded - that his industry had not delivered much better returns than the S&P500. Now, this was earlier in the year. The recent stock market crash may have altered the sums, but it's unlikely to have made a huge difference to the attractiveness of venture capital. And the important part is that they weren't making much money even when times were good. They are hardly likely to be piling in when the conditions are bad.