The company's latest price cuts and executive reshuffling prompt the questions: Is the architectural war between the PC and the Mac over? Has Apple lost and the Wintel platform won? More imp ortant, are the possibilities for further architectural innovations foreclosed by the dominance of the X86/Windows standard?

From where this observer sits, the answers to the first two questions are yes and yes. However, the third response is a surprising -- and resounding -- no. Let's take these issues in order.

It's hard not to look at Apple and wonder what might have been. Despite its numerous market missteps, the Cupertino, Calif., company still has a treasure trove of software technology. Its crown jewels include the Quicktime multimedia tools, personal-digital-assistant code stemming from the Newton project and the OpenDoc software components. Apple also has access to object-oriented software developed by the long-defunct Taligent and Kaleida ventures.

If Apple had played its cards right, it could have built upon such technolo gical leadership and secured its future.

Going forward, Apple still has two big irons in the fire. The most promising is QuickDraw 3D, which is being marketed to multimedia developers as an attractive alternative to Microsoft's more cumbersome Direct3D application programming interface. QuickDraw now runs under Windows, so developers owe it to themselves to check it out.

The second -- and far shakier -- effort is the construction of the Rhapsody operating system. Apple bailed out of Copland, its way-late successor to the MacOS. Now, the company says it will build Rhapsody on the Mach microkernel originally developed at Carnegie Mellon University and the OpenStep object-oriented components written by Next Software Inc., which Apple acquired from Steve Jobs in December for $400 million.

Mach is heavy-hitting technology. It supports multiprocessing, the execution of multiple threads within a task, and message-based interprocess communication. However, it's anybody's guess whether Apple will be able to deliver Rhapsody in a timely fashion and whether the new OS will be worth the wait.

The other questions I posed at the top of this column are whether the Wintel architecture has won the platform wars convincingly and decisively, and whether the possibilities for further architectural innovations are therefore closed by the overarching dominance of the X86/Windows standard.

Clearly, Wintel has won. But just because it's dominant doesn't mean computer architects aren't floating alternatives. Lots of people own Chevrolets, but that doesn't mean they aren't itching to try something a little more fun to drive.

The proof of the pudding is that the guardians of the Wintel platform themselves aren't standing still -- they're running as fast as they can to stay ahead of the pack. Indee d, big architectural changes are taking place right under the PC designers' noses. The biggest platform news of last year was Microsoft's launch of its Simply Interactive PC specification, a platform that combines MPEG multimedia, low-latency audio, digital simultaneous voice and data modems, and DirectX. The thousands of details Microsoft wants OEMs to fold into SIPC-like systems are officially documented in the PC97 hardware design guide.

In 1997, multimedia computing on the Wintel platform is also being defined by the MPC3 spec, which is the latest update to the Multimedia PC standard that's now several years old.

But if X86 proponents are firing off their big guns, other hardware designers are sharpening their shivs.

The most interesting Wintel alternative is the network computer -- the $500, stripped-down Internet browser that 's being touted by Oracle chief Larry Ellison and Sun CEO Scott McNealy. Oracle has established a subsidiary called Network Computer Inc. to move the NC architecture to market. (Britain's Acorn Computer is the company that will build a low-cost NC client box.) Sun Microsystems is hoping to spawn low-cost NC clients with its new line of Java-specific microprocessors. IBM has also set forth an NC reference architecture. Although systems aren't yet selling in force, there's so much activity it's hard to tell the players without a scorecard. To keep up, I check in periodically with the Network Computer News Service.

Another interesting architecture, the BeBox from Be Inc., never caught fire. It creaked along for a couple of years without gaining any market support. The BeBox was a low-cost multiprocessing system conceived by former Apple executive Jean -Louis Gassee. Be Inc. is still around, but it is now concentrating exclusively on its BeOS software, positioned as a platform on which digital content can be created.

Even mainstream Wintel OEMs are looking for an edge. Gateway 2000 is reshaping the traditional PC into the Destination Big Screen PC, a high-end box hooked up to a 31-inch TV monitor. RCA and Compaq are working together on a combined TV and PC. And Zenith is working with Oracle's Network Computer spinoff on a family of "Internet TVs."

All this activity indicates a big change is in the offing. During the next two years, a thousand flowers will bloom, as every vendor from Dell to IBM put its own stamp on the PC architecture. All this activity spells the beginning of the end of the one-size-fits-all personal computer, with which we've lived and grown up over the past decade and a half.

Have an opinion, comment or question about the Wintel platform? Explore next-generation design challenges -- things like MMX, DirectX and Univeral Serial Bus -- with other EE Times readers on our new Wintel Watch forum.

Free Subscription to EE Times First Name Last Name Company Name Title Email address   Click here for your Free Subscription to EETimes Europe