Out with the Old, in With the... Improved

by Alan Zisman (c) 2000. First published in Canadian Computer Wholesaler, January 2000

If you?re reading this column, I can assume that we all made it safely through the dread Y2K New Year?s, and that we can all breath a sigh of relief, and see what we have to look forward to. Some things I expect we?ll be seeing this year:

-- Computers will continue to get cheaper. Late in 1999, for example, Intel announced its ?Timna? chip, the company?s latest offering for the low-priced market. Timna will package a Pentium-class processor, though optimized using the latest 0.18 micron manufacturing. As well, the single chip will also include a graphics engine and a memory controller, combining three core functions in one. This results in smaller motherboards, and lower overall costs.

Timna?s all-in-one strategy is reminiscent of CPUs from Intel-competitor Cyrix. While Cyrix hasn?t been much of a force in the low-end marketplace lately, we may be seeing its return. The company has been purchased by Taiwan?s VIA, who have also bought up Intel-competitor Centaur. They have announced that they will continue Cyrix?s MII line, enhancing it at least to the 433 MHz equivalent for Pentium-style Socket 7 motherboard designs.

VIA expects to follow that up with processors still using the MII technology, but built for the newer Socket 370 standard, with equivalent speeds up to 566 MHz. As well, they plan to use Centaur?s WinChip technology to built Socket 370 processors with speeds of 500 MHz and up.

So we can expect increasingly powerful computers as Intel and VIA duke it out for low-end market share.

-- On the higher (and more profitable) end, CPU speeds will continue to rise. At the end of 1998, Pentium-II speeds had topped off at 450 MHz. By the end of 1999, high-end processor speeds were over 700 MHz, with models from both Intel (P-III) and AMD (Athlon). Both companies will continue to try to one-up one another throughout the year?assuming AMD can keep from bleeding to death on red ink.

I wouldn?t be surprised by gigahertz processor speeds before year?s end, with mid-market machines cruising at 700-750 Mhz.

-- Rambus memory-based systems are finally out, as Intel has released its much delayed 820 chipset. Despite the hype, initial reactions are that it offers only marginally improved performance over other, less-expensive memory models. As with the Intel vs AMD CPU race, but I don?t expect a clear-cut winner to emerge this year.

-- Expect Intel to release its 64-bit CPU, recently given the brand name ?Itanium? replacing the development code-name Merced. (Where do they get these names?) Expect virtually no impact from this oddly-named processor this year, especially as its 32-bit performance will be behind that of high-end 32-bit CPUs. AMD is promising its own, incompatible 64-bit design, code-named Sledgehammer, which it claims will be more compatible with 32-bit code than Itanium.

-- Expect to see more use of USB to connect printers, scanners, cameras, and other relatively low-bandwidth devices to both Macs and PCs. But also expect increased grumpiness with USB as users try to use it for higher bandwidth devices?network adapters, cable and ADSL modems, speakers, and video and more.
Whether named FireWire (and Apple trademark), the awkward IEEE1394 or High Performance Serial Bus (HPSB), I suspect this higher bandwidth bus solution will fail to widely catch on with the PC platform, though it will be built into virtually all Macs. PC manufacturers will, instead, wait to see what happens with the higher-speed USB-2 spec.

Meanwhile, I?d expect to see the 1984-generation ISA bus linger on a bit longer, complete with legacy parallel, serial, and PS/2 ports, continuing to plague efforts to make the PC platform reliably plug and play. Ideally, by this time next year, it will be finally dead and buried?but I wouldn?t be surprised if the burden of backward compatibility and the conservatism of the PC platform keeps the ISA bus limping on even longer.

-- Expect PC makers to try out more innovative designs, as they finally manage to break the stranglehold of the beige tower design, while avoiding copying the iMac look and feel. But expect the bulk of the market to remain boring beige boxes. (Come on guys! You can do better!)

-- Microsoft will release Windows 2000 as promised on February 14th, but while it will end up on systems destined for the corporate market, it will have a minimal impact on the home and small business market. Microsoft will continue testing its so-called Millennium operating system, its proposed replacement for Windows 98. Will it be released this year? No bets.

Linux will continue to grow. New releases such as Corel?s will make it easier then ever to set up, configure, and use a Linux system. As the judge suggested in the Microsoft judgment, however, Linux will remain a niche market, with more impact as a server operating system than as a mass desktop OS. (And the real geeks will spurn Linux, as ?too commercial?, looking at something like Free BSD, instead).

Apple will release OS X, confusing its current devoted user base with this break with its legacy code and applications, but will manage the transition more smoothly than could ever happen on the PC platform. With growth to close to 10% of the market share, Apple will be back where it was 10 years ago?an interesting alternative to mainstream computing, that can provide a comfortable and profitable market for companies prepared to be big fish in small ponds.

-- Thin clients will be modestly successful in the big business market, but ASPs, so-called Application Service Providers will fail to catch on in a big way. The idea of storing applications and data across the Internet is attractive to many, but it?s too dependent on a still-fragile network to be something that either small or big business can rely on.

-- We will see an increasing number of small computer-like devices, with the Palm OS continuing to steamroll Microsoft?s CE. There will be more use of wireless connections, both for local area networking (Apple?s AirPort has a lot of promise), and for accessing information across the Internet. But both small devices and wireless connections won?t really become pervasive for another year or more.

All in all, a transistio