The Next 25 Years . . .

Add Your Comments

It has been twenty-five years since IBM launched the IBM PC, and anyone who doesn’t think that the PC (along with its companion, the Internet) hasn’t been one of the top five historical events in the last 25 years is probably old and definitely computerphobic.

If you want to look at some PC history and track the growth of the industry, you could do worse than going here, and here.

If you’re more interested in the next rather than past 25 years, what do we think the PC world will look like in 2031?

We think it will vanish.

Oh, there will be a relative handful of PC-looking boxes here and there doing all sorts of interesting things, but relatively few people will only them, and little else will be the same.

We don’t expect to see AMD around in 2031. For that matter, we don’t expect to see Intel, either. Or Microsoft. Or Dell. Or HP. Or most of the current players, at least not in anything resembling their current form.

What do we expect to see by 2031?

Embedded Everything

What we expect to see by 2031 is that anything and everything that could use some computing will have computing built-in.

That’s certainly common today, the difference will be that those little embedded systems will be (when needed) the equivalent to today’s big box, or more.

The PC will not become the commander of the home entertainment system, the home entertainment system is going to eat the PC. Or really, PCs.

The general purpose processor will become a beast from the past. What you’ll have instead are simpler, dedicated processors (probably clumps of them), specialized to handle just a few tasks very, very well. And, if you have more than one task, you get more than one clump.

If you can make the cost of a CPU nominal, there’s no more point having a do-it-all processor than a do-it-all set of clothes. Have a clump for office work and another clump for media, and put or plug one or both into your screen or phone or both, just like you wear different clothes for different occasions and climates.

For instance, by 2031, we think games will no longer be passive code acted upon by some outside agent, but rather little CPUs themselves, a data and processing cube telling the rest of whatever else is plugged in what to do.

This may begin as a way to end illegal copying, but not for long, because we think by 2031, people will be able to make their own computer chips with little more difficulty than making a copy of a movie.

By 2031, we will see a revolution in fabrication. It will just cost too much money to build the fabs to build zillions of CPUs the current way by then. Instead, we will see less-than-cutting-edge but very cheap techniques build generic chips that can be easily programmed to whatever function might be desired. Imagine EEPROMs on supersteroids.

Initially, those doing this will probably go by the names of Samsung and the like, a bit later by software manufacturers sending out both programming and fabrication code via the Internet, but by 2031, I think at least enthusiasts will be building their own chips, too.

Speaking of the Internet, communications will be much faster and will be every place with at least a hint of civilization. The thought of being tied up or evenly seriously encumbered by either computer or communications will seem . . . savage.

The Price To Pay For This?

Today’s latest processors are arguably thousands of times faster than the 8088 processor that powered the original IBM PC. The CPU of 2031 won’t be.

Oh, it will be faster, all right, just nowhere near a thousand times faster, and most of the speed increase will come from simplication, specialization and distribution of computing tasks, not brute force frequency increases.

Next-generation computing will be about flexibility, specialization, efficiency (energy and otherwise) and mobility far more than brute speed.

In time, brute speed measurements will wither away. By 2031, your next generation, no doubt for a school project, will look up “FPS” in an encyclopedia (probably some son of Wikipedia) and read/hear that FPS was “a primitive means of measuring video card performance.” The next-generation video system will be capable of far more complex and realistic actions; it just won’t deliver them any faster than screens can handle them. Overkill FPS will be a tossed frill.

Death of the Dinos

You may say, “OK, even assuming all you’ve said turns out to be true, why do you think all the leading computing companies are going to die like dinos after an asteroid strike?”

The answer to that is simple: these companies lived by the PC and are going to die with it. These companies have lived inside the PC box, and they really can’t think of a world without one.

There’s a mindset associated with the PC: a large, complex, general purpose machine: difficult and expensive (but very profitable) to build and feed, complicated to use.

In contrast, dedicated specialized circuitry will increasingly be able to get the (relatively modest) real job done with much less fuss and expense and size, and people will vote with their wallets.

And, believe it or not, most people will be happy to say “Good riddance” to that big ugly box once they’re given the chance to.

However, inserting tiny little computers into every little geegaw will destroy the PC world: its assumptions, its economics. Intel can’t charge hundreds of dollars for a computer in a hundred dollar phone; Microsoft can’t charge a hundred or two for an OS in the same phone, and Dell can’t charge %500 to sell them the phone.

They’re likely to fight off the extinction of that very profitable world with everything they have, first by trying to keep the big box king, then by trying to turn future phones into $1,000 items.

And, with no doubt a few exceptions, they’ll probably die trying.

Ed


Leave a Reply

Your email address will not be published. Required fields are marked *