From the very first days of the PC revolution, the ever-increasing sophistication of software has prompted users to demand faster and more powerful hardware. As fast as Intel could bang out a new chip, software developers would hit back with "next-generation" applications that soaked up all of that processing power and still demanded more.
It's been a pretty good business model, especially for Intel. But it's also been great for every reseller and integrator too, who have been virtually guaranteed that about three years later, the customer would be back for an entire new system. Imagine how happy a retailer of furniture or white goods would be with that upgrade cycle!
Do you get the feeling, though, that this model is starting to show signs that it is not sustainable? Is it just my perception, or are new software developments starting to lag behind the hardware? As ever, Intel is adhering to Moore's law by doubling processor speed every year, and last week announced its road map for the next line of processors (page 6). But do we really need them?
I'm tapping this editorial out on a Pentium 200MHz MMX Toshiba notebook. I run the latest version of Microsoft Office and Lotus Notes, which is about as resource-hungry as any desktop application out there. And I'm happy.
I absolutely have no desire to get a faster machine. I'd probably take another 32MB of RAM and a larger hard drive, but in terms of processor speed I'm perfectly content. I'm satisfied with a machine that would be lucky to be considered entry level.
Not particularly taxing
Why? Because the applications I use don't particularly tax my system. What's more, I'm pretty happy with these applications. They could be a bit more stable, but in terms of features and functionality, again, I'm pretty satisfied. I'm certainly not hanging out for Office 2000 or Notes 5.0.
I wasn't surprised, then, when I read Tom Allen's front page report this week on the burgeoning business of "remarketing" second-hand computers. Quite simply, most people today are going to get by with these lower spec'd machines. Perhaps the only thing stopping this market from exploding is that for so long it has been drummed into users that they need to buy powerful machines, otherwise they will be good for nothing within a year or two. I don't think this is the case anymore. But I promise not to tell customers, if you don't.
Of course, this could be a short window of opportunity. All it will take is another burst in software innovation and we'll all be crying out for new machines again.
People often point to the latest games as pushing performance boundaries. However, even they are adequately playable on slower machines, even if the graphics aren't quite as pretty or the scrolling quite as slick.
If I had to point to anywhere where the "next generation" of applications might spring from, I'd point to voice recognition, which seems to be on the verge of being truly useful today (see page 18).
Until then, though, I'll just keep tapping away on this little notebook, with a smile on my face.