NEW YORK (CNN/Money) - Tech investors (and there are a few of them left) are obsessed with pinpointing the elusive bottom -- that magical moment of maximum pessimism after which tech stocks can only go up.

Many (myself included) think we're not likely to get there until we endure a scary, cathartic capitulation, a giant frenzied selloff that frightens all but the most committed out of tech stocks.

But what if all this talk of "bottoms" and "capitulation" is actually blinding us to decent buys in the tech world right now? That's more or less the argument of Soundview tech strategist Arnie Berman, who argues that tech stocks finally have fallen far enough to be a "great buy" at current prices.

Tech investors, Berman observed in a recent research note, are currently "depressed, lethargic, discouraged and afraid to buy stocks." He argues that "investors are far more worried about the market than they should be," with even those "tempted to buy dissuaded by the fear of being 'early.'"

Berman himself has to plead guilty to the "early" charge: He's been pounding the table pretty hard since May. But there are worse crimes than being early. Berman argues that you can make good money buying reasonably priced stocks somewhere in the general vicinity of a bottom.

Berman notes that valuations in the tech sector have come down precipitously in the past several months. The typical tech stock, he notes, has dropped roughly 30 percent since mid-May -- with most chip stocks falling even further than that. Yet he doesn't think the future of the tech sector is "nearly as dire as technology stock prices currently reflect."

MORE TECH INVESTOR       Intel's big problem Yahoo's tough transition What the contrarians see

So what does Berman suggest investors buy? Citing recent good news from Dell as evidence that "PC demand conditions are not nearly as bad as the consensus fears," Berman actually thinks some PC-related stocks -- including Dell, Intel and Microsoft -- look ripe for the picking.

Many tech observers expect software to see the biggest boost when tech spending picks up again. But Soundview's surveys of corporate infotech departments suggest that software isn't what most business managers say they really need. Once companies move away from the current focus on "not spending," Berman argues, companies that have been skimping on hardware are likely to move more quickly to replace rapidly aging PCs, many of them bought in advance of Y2K.

Those in charge of writing the checks for IT tell Soundview that "the first checks they will write will land in the pockets of their big hardware suppliers. In a period where nearly every form of postponable spending has been postponed, most users have found it easy to delay the purchase of the next PC, the next mobile device, the next mainframe, the next storage device." But they won't delay these essential purchases forever.

For a decidedly different take, see my last column, on Intel, which elicited strong reactions. My argument went like this: Intel's chips keep getting better -- but that doesn't help it unless people want to pay up for those better chips. And they won't, I wrote, because the chips they already have can handle most of their needs right now and in the near future.

Some agreed that there was something of a "performance glut" hanging over the PC sector. Others thought I was talking like a Luddite. "Business always has a need for faster, more efficient PC's for their employees," wrote one. "I have a spare 300MHz PC in my guest room...do you want it?"

An unintentionally ironic question, given that I do much of my writing and Internetting on a four-year old iMac with a 266MHz processor. (Sure, I have a much newer and zippier laptop, but, believe it or not, the old machine works fine for most purposes.)

Still, there's no question that someone will come up with a new killer app that can truly takes advantage of the processing power that Intel can offer. What that will be, and when that will happen, I don't know. One reader suggested that speech recognition -- which demands massive juice -- might do the trick, though speech recognition is still so glitchy, and has been hyped so hard for so long, that it's a little difficult to get excited about its prospects, at least in the near term.

I'd be interested in hearing what other readers think might drive demand for processing power in the future.

Sign up to receive the Tech Investor column by e-mail.

Plus, see more tech commentary and get the latest tech news.