April 24, 2003

Little machines can make a big difference

Posted by Scott at 10:25 AM

Andy Ihntako of the Chicago Sun Times writes about how little we actually need all the processing power of the high end computers.

"Just how important is raw processing speed? Not very, it turns out.
Chasing after the most powerful PC in the store can be a fool's errand. You're spending money on processing cycles that you'll never use when, instead, you could be spending it on actual useful features, like more storage or wireless networking. "

I remember one day when a co-worker saw me working on my (then) new iBook. He came by and said So, how many megahertz is that? It was his opening question. Never mind that the machine ran unix (which I prefer to Windows), had built in wireless networking, had 640 megabytes of RAM, turned on and off as easily as a Palm Pilot, or had ~4-5 hours battery life. No, all he wanted to know (perhaps to reassure himself) was how many megahertz the laptop was. It's a laptop for goodness sake! Not a compute server!

The rest of the article makes some rather good points. I recommend it if you're having trouble thinking outside the multi-gigihertz PC tower mindset. He makes some interesting references to the state of cell phones in Japan as well.

"Maybe they [the Japanese] had the right idea all along: when your access to the Net comes via a cell phone, you have a simple and non-intimidating device that provides you with information, e-mail and digital services. You get it by walking into a retail shop, plunking down about $100 and agreeing to pay a low monthly fee for access and service. It's with you all the time and most importantly, it always works.
Compare this with the way you get the "real" Internet in the U.S. For $700, you get a bare-bones PC that requires regular maintenance and runs complicated, fiddly sets of software that don't play nice with each other. Access to the Net will still cost you about $25 or $30 a month, and you can't simply stick it in your pocket and take it with you. "