Computers 2039

Add Your Comments

There is an allegedly Chinese curse which says, “May you live in interesting times.”

nVidia had an “interesting time” at the stock market yesterday. After it said it was going to cost them $150-200 million to fix some problems they had with notebook graphics chips, and that they weren’t going to be making as much money as they had expected in the near future, the stock price dropped 30%.

Assuming you’re not a nVidia stockholder, should this concern you?

No.

For those who want to buy a company’s product rather than its stock, there are two kinds of bad financial news. There is bad financial news that seriously hurts the company’s ability to develop and make good products, and there is bad financial news that doesn’t.

Customers should be concerned about the first, but need not be too concerned about the second.

nVidia’s problems fall into category two. The company’s second quarter profit will go to the notebook chipfix, and they won’t make the kind of money they had expected, but the company has no real debt and does have about $1.8 billion in its piggybank, so these problems are by no means crippling.

This doesn’t mean you should sell all your have and buy nVidia stock because the company is faced with an era of “interesting times,” potentially crippling, life-threatening long-term challenges that makes today’s problems seem trifling.

In the long run, nVidia is threatened by its most profitable market, video cards, dwindling away, and being frozen out of the integrated video market.

You may say, “Video cards dwindling away? What about gamers?” That’s the problem, video is getting to the point where separate video cards will be bought pretty much only by gamers.

That’s bad enough, but the real problem is in integrated video. In a few years, the typical vanilla CPU and especially tinyCPU will have free video built into it. Who needs an extra video chip?

Of course, some people will still want video cards and some people will still want integrated video than can be squeezed into a CPU, but these trends are definitely a much bigger threat than nVidia has seen in the past.

As we’ve mentioned before, this is why nVidia is coming up with things like Tegra.

nVidia’s problem with getting into the CPU field is that while they are financially solid, they’re also financially small. By revenue, nVidia gets roughly the same revenues that AMD gets from its CPUs, and is rather smaller than Intel’s chipset/mobo business.

nVidia is a profitable company because the graphics chips business has been able to duck most of the huge capital expenses associated with chip fabrication (especially bleeding-edge chip fabbing), while still being able to charge enough to more than cover the additional production costs of going fabless. Compared to the CPU business, the graphics business has been downright cozy. Size didn’t terribly Matter.

In a few years, though, the competitive market will likely look quite different. If the economics and means of providing graphics becomes that of the CPU industry and Size starts Mattering a lot, nVidia is in a lot of trouble.

It’s important to understand why Intel’s Larrabee is so dangerous to nVidia. It’s not just a matter of a much bigger company launching a video card, bad enough as that is.

No, Larrabee is just a part of Intel’s vision of the future, which is “Our strategy for the new world coming might as well be called the Swarm. We throw a swarm of tinyCPUs at any and all computing situation. Small situation, small swarm. Big situation, big swarm. And since we can make more of them, and make them better and cheaper than anybody in the world, we’ll rule the new world even more thoroughly than we did the old one.”

It’s a strategy meant to maximize Intel’s advantages in fabbing and economies of scale to the max, and if it works, if a swarm of tinyCPUs can do the job as well and cheaper than one big expensive custom-purposed chip, nVidia (and a lot of others) are roadkill. It’s Intel turning into Walmart, and everybody else becomes the local store.

This doesn’t mean nVidia is doomed; Intel could well flop. But Intel’s Swarm Strategy represents a potential threat to nVidia far, far greater than anything it has faced before.

AMD? Well, if they’re around that long (and unless they get bought, or get unbelievable amounts of money from their lawsuit, they won’t), they’ll be doing the same thing.

None of this is going to have any immediate or short-term impact, and nVidia certainly has a number of big plusses on its side. But there’s little doubt the next five years will see nVidia move from a relatively comfortable competitive environment into a far more hostile one.

The same can be said for the PC industry in general. Yes, we may well be heading into some rough economic times for a while and the PC industry will take its lumps, but a much bigger, more fundamental change is beginning to happen. It’s not going to be more of the same. The entire PC industry is going to go through a radical change, just part of it being making many more units for much less money. A lot of today’s PC giants aren’t going to be able to make the transition.

This isn’t just a matter of nVidia’s future being insecure. Nobody big in the PC industry today is secure.

Ed


Leave a Reply

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