Thanks to strong global mobile PC sales, Intel hauled in $1.6 billion in earnings on $9.5 billion in revenue for the quarter that ended June 28, better than recession-wary analysts expected on average. The company’s projections for the third quarter were upbeat, too: sales as high as $10.6 billion, better than the pundits had guessed, as consumers and businesses are expected to snap up chips in computers from Hewlett-Packard , Dell, , Apple and others. Nonetheless, the stock ticked upward only about 1 percent in after-hours trading.

With all that good news, it’s fair to ask why the stock barely budged. And while it’s often tough to pin down one reason, this time there’s a likely culprit: the company’s flagging profit margins.

Intel, being the biggest manufacturer of computer brains on the planet, typically commands a hefty premium for its wares, but this quarter it had to work a bit harder for the money. The company’s gross margins (one measure of profitability) came in at 55.4%, a tad under management’s projection of 56%. Doesn’t seem like a big deal, right? Well, when you’re dealing with billions of dollars in revenue, every decimal point counts. So while Intel analysts seemed generally pleased by the results, on a conference call with CEO Paul Otellini and his team Tuesday afternoon they peppered him with questions about that missing .6% of margin. Where did it go? Why? Will it be back?

The answer from Intel executives: Laptop-hungry consumers sank the margins, because they’re ditching desktops faster than expected. And no, the margins won’t exactly be coming back as strong as they once were.

Consider it the slight downside to the world’s blossoming love affair with the laptop. Though folks at home are increasingly buying them to watch video, surf the web and digitally socialize, the surge in demand from price-conscious consumers means Intel ends up selling more low-end, lower-profit laptop chips. This trend could spell trouble for Intel’s long-term profitability, unless the company can figure out how to lower its production costs.

Of course, that’s exactly what Intel is trying to do. In the chip game, the way to lower production costs is to crank up production, because while it can cost billions of dollars to invent a chip and build a plant to manufacture it, the raw materials involved are pretty cheap. Those economics mean that it costs a lot more to produce the first chip than the next 100 million.

That’s one reason why the company’s Atom chip is important. Atom, launched in March, is Intel’s smallest chip, intended to power wireless devices and stripped-down PCs. It’s not powerful enough for a great experience editing photos or viewing online video – but Intel designed it as a cheap way to get more people connected to the PC-based Internet. It’s a clever Trojan horse strategy to get PC power into the hands of consumers in developing markets, and it seems to be working.  Intel executives say Atom chips are selling five or six times better than they projected last year. Atom’s gross margins are also slimmer than Intel’s overall margins, but that’s a hit the company is prepared to take in service of its strategy. It’s like Toyota getting first-time car buyers to go home in a Yaris, hoping they’ll one day upgrade to a Lexus.

In this case, the upgrade from Atom is the Centrino 2 platform, which the Intel unveiled on Monday. Centrino 2, a combination of a processor, chipset and wireless functions working together, is Intel’s mobile tour de force. It plays and decodes high-definition movies, conserves power, and thanks to three wireless antennas handles 802.11n wireless signals at a top speed of 450 megabits per second. (Which, you’ll have to trust, is very fast.)

Centrino 2 has just begun shipping in volume, and it’s part of the reason why the company feels good about projecting healthy results for the third quarter. Executives had just better hope that in this Yaris economy, there are still enough laptop buyers out there with Lexus tastes.