In this holiday season, I am comforted to read that Moore's Law, which only recently seemed headed for repeal, now appears safe for another decade or two. It's hard for me to imagine a transistor the size of three atoms, but Intel says it can build them and that's what counts, I suppose.
But what counts and what matters is different. And what matters is this: What has Moore's Law done for us lately?
Don't get me wrong, I am all for faster processors. But right now I need megabits much worse than I need megahertz. The problem I have isn't that my computer can't process data quickly enough -- it's that data doesn't move around the network quickly enough to use all that computing power.
And ever-teensier transistors -- the core of Intel co-founder Gordon Moore's observation that the number of transistors on a chip doubles every 18 to 24 months -- won't solve the problem by themselves.
Most people already have more megahertz on their desktops than they can really use. The slow sales of this year's Christmas computers seem to bear that out: Lacking a killer application that demands more CPU horsepower, people seem content with what they already own.
And that killer application hasn't appeared because most people, while they have the megahertz, don't have the multi-megabit Internet connections they need to make the applications usable. And for all the talk, those connections are years away for most Americans.
Intel knows this and has been grappling with the dilemma for several years. Nobody works much harder than Intel on dreaming up new ways for people to use processing power. But even as they build atomic-sized transistors, dreaming up applications has been a problem.
The big things I've seen Intel develop are, alas, media applications. And media apps hog more than just processing power; they also require a lot of network bandwidth that most people don't yet have.
I say "yet" because I think this is just a speed bump for the importance of Mr. Moore's law. As bandwidth increases, people will find cool things to do with it -- mostly entertainment and communications -- and applications will catch up to the power available to them. Then we will need new computers, and all will be right with the hardware companies again.
But like I said, huge bandwidth is, forgive the pun, just a pipe dream for most people. And the broadband we really need is bigger than what's even being offered today, so add some more time to the equation. Remember how long it took for "everyone" to get cable television?
So while Moore's Law seems to still have life in it, it would be OK with me if it took a little rest while bandwidth catches up.
ZDNet News commentator David Coursey is based in Silicon Valley and has covered personal computers, software, and the Internet for more than 20 years. He is an industry analyst and creator of several industry conference events. His Web site is www.coursey.com.