It took the market about 20 minutes on Friday to figure out that
midquarter warning was bad news for
Advanced Micro Devices
AMD, which has been stealing market share for more than a year, opened at $41.33 -- 41 cents better than the previous close -- after Intel released the bad news, indicating that investors figured that the smaller company had tweaked the giant's nose yet again.
Don't believe it, says Dan Niles, CEO of Neuberger Berman Technology Management. "Our belief is that AMD is doing well, but not that well. This indicates the PC market has some issues," he says.
Judging by AMD's subsequent swoon -- it closed down $1.83, or 4.4%, to $39.50 -- plenty of investors saw it the same way.
Simply put, AMD couldn't possibly have gained enough share to force Intel to drop the midpoint of its guidance from $9.4 billion to $8.9 billion, says Niles, whose company has a position in AMD but not in Intel. "The math doesn't work, since Intel's share is nearly 80%," he says.
Shane Rau, a semiconductor analyst with IDC, notes that AMD gained more than a point of share at Intel's expense in the third quarter and may have gained as much as another point in the fourth quarter. "But this is more a statement about end demand than market-share direction," he said.
Right now, it looks like semiconductor makers not tied to PC sales are likely to remain in good shape, particularly those selling chips to handset makers. We'll get a much better reading when
delivers its midquarter update on Monday.
Is Oracle a Google Wannabe?
announced that it developed a new search technology this week, much of the financial and trade press immediately assumed that Larry Ellison and company were moving against
Google's major Web and desktop search products are geared toward consumers -- and that's not a market Oracle is likely to pursue. Instead, the database giant is aiming at the $1 billion enterprise search market, which is now dominated by
and a host of best-of-breed players such as
Unlike simple Web or desktop searches, enterprise searches trawl corporate databases, emails and even unstructured sources such as help-center call logs.
Security is a big issue as well. Companies can't allow an employee who may have permission to search a database of component suppliers, for example, to also look at sensitive personnel records. "Security is central to what Oracle does as a database vendor. A secure search is a natural extension for them," says Sue Aldrich, an analyst with the Patricia Seybold Group.
Google does have an enterprise search device, but its capabilities are rather limited, she says, and it is by no means a dominating force in the market. Moreover, its enterprise search business is small, perhaps $50 million, and is aimed largely at the small- and medium-sized business markets, says analyst Jim Friedland of SG Cowen.
IBM, which partners with scores of smaller application providers, sells middleware that acts as a platform for very sophisticated searches. A researcher in the pharmaceutical industry, for example, may need information about a "BIKE." There are bikes that people pedal and BIKEs that are parts of a gene -- a distinction that a simple keyword search would miss.
Indeed, developing industry-specific search technologies is increasingly important, as is the ability to plug into other specialized search applications. In January, IBM started giving away an open-source middleware component that will make it possible for companies to combine search technologies from multiple vendors into their software stack with a minimum of expensive custom programming.
Also overlooked in the wake of Oracle's announcement from Tokyo was this: "The business intelligence companies like
( HYSL) and
( BOBJ) have really dropped the ball," says Rob Tholemeier, a former sell-side analyst who now invests broadly in business-intelligence stocks.
Tholemeier says the future of corporate search lies in mining data from unstructured sources, an area that the typical BI vendors have yet to emphasize.
Those teases at
have been dropping a lot of hints about "Origami," the mystery (choose one) a.) iPod killer b.) mobile PC c.) software platform d.) can opener, which is now due to be unveiled next week at CeBit, the big German technology showcase.
Mary Jo Foley, the astute Microsoft watcher and
journalist, says: "The software giant intends to bring to market a new class of smaller and more versatile, battery-powered computers that offer users greater utility -- they could connect to networks and download email or make VOIP (voice-over-IP) phone calls, as well as manage photos, music and videos -- but that still use a full version of Windows."
Oragami, she writes, will be smaller than the smallest Tablet PCs but bigger than a PocketPC and unlikely to ship with a keyboard. Instead, users will input data and commands with a stylus, according to sources quoted by Foley.
Microsoft could have a difficult time storming the store shelves with this one, says Gartner analyst Michael McGuire. "If it's an entertainment-oriented device, the iPod and other MP3 players have set expectations very high," he says. It will have to be small enough to fit in a pocket (Windows-based devices tend to be bulky) and be priced competitively.
Another problem: Lots of folks in the tech-savvy audience likely to be targeted already have wireless notebooks and handheld devices of one kind or another at home and at work, so Origami had better be good.