CES Strains Seams, Credulity

After two years of echoing emptiness at Comdex Fall, the bustling crowds at January's Consumer Electronics Show, straining the seams of the Las Vegas Convention Center, were almost too much to handle. It was a great show for microprocessor developers; all the best products at CES are microprocessor controlled, and many strain the limits of processor technology.

Microsoft chairman Bill Gates announced a substantial expansion of the company's Media Center PC concept, with three new products designed to allow the sharing of digital content throughout a home. A new low-cost Media Center Extender set-top box will use Microsoft's Remote Desktop Protocol (RDP) as a remote display device for any Media Center PC on the local network, providing the same Media Center user interface as the PC but without the need for local storage. The Media Center Extender can be much smaller and cheaper than a PC, so customers can afford to buy one for every TV in the house—at least, that's Microsoft's hope.

For those who already have a Microsoft Xbox, or who might be willing to buy one for this purpose, Microsoft will offer a Media Center Extender kit, consisting of a remote control and a DVD with the RDP software. Like the set-top box, an Xbox running this software will provide the Media Center PC user interface, but at a much lower cost per TV.

Finally, those who want their video to go will be able to get a Portable Media Center. Gates showed a prototype model from Creative. Though bulky, presumably because it needs a large battery to achieve reasonable operating times, the Zen Portable Media Center may do better on the market than current products that lack Microsoft's support. Creative plans to ship the device in 2H04 at a price of $499 with a 20G hard disk.

Some products at CES strained our credulity, especially those in the home and car audio markets. Fortunately, there were no eardrum-straining indoor demonstrations of just how powerful today's subwoofers are, but many companies were offering models that could easily double as lithotripsy machines for breaking up kidney stones without surgery.

What bothered me more than any of these products were the ridiculous claims of vendors of high-end interconnect devices. (We define "interconnect" as an audio or video cable that costs far more than it's worth.) At least one company at CES was showing audio cables with a battery holder and an extra center conductor to apply a DC electric field across the dielectric to "align" the dielectric molecules, thus "smoothing" the sound. As an engineer, I believe in impedance matching and full-coverage shields, because these things are easily proved to matter. Dielectric molecular alignment, on the other hand, is simply a fraud. The best one can say about it is that it doesn't hurt the electrical properties of the cable.

It gets worse, however. The Sony/Philips Digital Interface (S/PDIF) standard, which defines electrical and optical interfaces for digital audio, dates back to the 1980s. Optical S/PDIF interfaces often follow Toshiba's Toslink specification for connectors and cable, which was designed to provide error-free operation using inexpensive plastic fiber with a 1.0mm core diameter. That's simple enough, but I saw a company at CES demonstrating Toslink cables constructed from a fiber bundle—that is, multiple independent fibers. In this case, the bundle was made to be coherent; the relative positions of each fiber were maintained from one end to the other, so the cable could actually transfer low-resolution images from one end to another, just like a medical endoscope. This technology reduces the functionality of the cable, because the bundle will capture and transmit less light than one large fiber.

I believe the high-end audio industry has developed these bizarre practices because audio technology long ago passed the point where ordinary customers could distinguish between new and old products. The tremendous improvements in sound quality achieved in the 1960s and 1970s paved the way for CD digital audio, but even that high standard was well integrated into low-cost products by the late 1980s. Diminishing returns on further investments in quality have driven the audio industry to a form of madness.

Now, here's the payoff. Are we in the microprocessor industry at risk for the same disorder? We, at least, have customers who genuinely need all the performance we can deliver, essentially without limit. That alone guarantees Microprocessor Report a long-term future. Some computer buyers, however, are already satisfied with the speed of their systems, and I don't know what we can offer these people to get them to upgrade. I just hope we won't be reduced to touting snake oil and silver bullets.