I spoke yesterday about AMD possibly being an early casualty of the “less is more” phenomenon.
That would be good for Intel, for a while. Not long.
Looking further down the road, it’s hard to see how the “big box” computer can continue its reign.
The outsides of the typical computer haven’t changed very much in twenty years. The insides certainly have, and the theme has always been “more, more, more.”
What happens to “more, more, more” when you reach “enough?”
In the dinosaur age, Tyrannosaurus Rex ruled, he was a big mean mother. But when times got bad, he wasn’t quick and nimble enough to compete against the little mammals.
We are having our own asteroid strike in the tech field.
Big Box or MP3 Player?
By 2012, I think the typical computer will look more like an MP3 player than the big boxes of today.
The typical computer of 2012 will be cheap. It’s pretty hard to see how a computer with essentially the same firepower as the best available today would cost more than $100, might be $50. It’s even harder to see why the average person would want to spend $1,000 for some megamonster box PC if he can get the equivalent of today’s best for $50-$100.
The typical computer of 2012 will be portable. They’ll fit into your pocket. They’ll use wireless for all communications (they’ll be phones, too). Outside of a relative few power users, there will no longer be such things as home or work computers, you’ll carry your computer with you whereever you go, and hook it up to a bigger LCD screen at home or work whenever you need one (the more serious will use notebooks, but the day of the foldable notebook will be coming).
The typical computer of 2012 will be specialized. You’ll have communications/entertainment/basic business computers. You’ll have gaming/entertainment computers (some no doubt will be built into HDTVs). They’ll be much simpler to use simply because they won’t be designed to do everything; just a few things well.
The typical computer of 2012 will be durable Many computers won’t come with storage. It won’t be necessary when you can have 50 gigabytes of so of nonvolatile memory instead. Even if that is a little too optimistic for 2012; tiny hard drives will be read electronically rather than mechanically.
In all likelihood, there won’t be a whole lot of need for storage because outside of some trifling personal files, you won’t have or need things like music files. You’ll pay for wireless music channels and movie channels, and the idea of actually owning songs and movies will no longer be cool, and on the way to becoming a little quaint.
The little boxes will have digital rights management built into them, and since the nuts and bolts of these machines will only be accessible to a few, most people will pay for their various wireless music and movie channel plans just like they do for wireless cell phones. Too much hassle to do otherwise.
The Internet itself will be relatively smaller. Again, cable will provide a model for this. You’ll subscribe to services, and most websites will get paid a small sum based on where you go. Privacy people will squawk a bit, but since there will be no need to track who goes where, (except maybe for premium sites), they will quiet down quickly.
Just like cable, paying a fee won’t mean no advertising (though it will be less obnoxious than what we see today).
The Big Box Niche
Big boxes won’t perish from this earth; they’ll just go into hiding.
Primarily, you’ll see them in servers. That’s where you’ll see things like 10 terabyte hard drives.
Fewer will serve for true power computing, for those relative few that really need it, and even most of those will probably be more likely to have multiple processors than very powerful processors, for a few reasons:
First, faster and faster processors are simply going to become harder and harder to manufacture cheaply. Silicon and its close relatives are reaching the end of their useful lives in the “cram more and more circuits into less and less space” development model. If nothing else, heat will become an overwhelming factor and solutions to it will become more and more expensive.
More importantly, though, the need for such devices will drop as the vast majority of users either find themselves with enough power, and most of the remainder find multiple processors the cheap path of least resistance.
This will be bad news for gamers. Essentially, Joes Sixpack and Suit have been partially and indirectly subsidizing gamers by continuing to buy more and more powerful machines. You can afford a $400 video card because the rest of the computer doesn’t cost you much more than that.
When the Joes stop doing that, gamers will be faced with either buying the gaming boxes built for Joe, or will be laying out thousands rather than hundreds for that 10GHz system. When 5% rather than 75% of computer buyers are interested in Pentium VIIs, the proud, the few, Pentium VII owners are going to have to pay those multi-billion fab costs now spread out over everybody.
After a while, very high-speed computing will become for a while the province of governments and especially the U.S. military as they spend the money and do the research to come with alternatives to silicon that will make what I call the Second Hardware Revolution possible. Just like they did the First.
The Computer Revolutions
The first computing revolution was “bring cheap computing to the average person.” That’s the PC.
The second computing revolution was “bring cheap world-wide communications to the average person.” That’s the Internet.
The third computing revolution is and will be “bring cheap media (i.e., sound and video) communications to the average person.” We don’t have a name for it yet.
The fourth computing revolution will be “bring cheap virtual reality to the average person.” That’s William Gibson land, and that’s probably 2025-2050.
The first two revolutions are almost finished in the developed world. In the next ten years, we’ll go from relatively cheap to practically free, so that most of the rest of the world can join in.
The third is just beginning, but it differs from the first two in that it really doesn’t need more powerful computing for it to happen. Present power is more than good enough.
The fourth is beyond the current computing paradigm, and will require hardware technologies we can only guess at now.
Cheap, Cheap, Cheap
The significance of the PC or the Internet is not their existence. You could do the equivalent of everything the PC or Internet can do before the two existed. The problem was how much it would cost you to do it back then.
It has been, is, and will be technological cheapness, not technological achievement, that makes these significant to the world.
Up to now, technology has been able to give us “more for no more.” Now that the industry has grown up and become a commodity industry, people are now beginning to look for “the same for less.”
Not just less money. Less hassle. Less inconvenience. Less complexity.
This is what most people want. This is why handhelds have become so popular, even though they are relative Tinker Toys compared to a notebook. This is why people carry around MP3 players rather than notebooks. This is why wireless is exploding.
But the technosaurs don’t want that. How can Intel charge $150 for the CPU in a $99 computer? How can it do so while spending $20 or $30 billion for a fab plant to make .01 micron chips that chew up 150 watts of power and require Peltier Plus cooling?
How can Microsoft do the same on the OS side?
Apple essentially destroyed itself in the early nineties because at a critical time, it couldn’t conceive of a sub-$5,000 good computer.
The people on top of the computer world today are likely to become tomorrow’s Apples.