PC architecture shift.
The announcements at the Microprocessor Forum from IDTI, AMD, Cyrix and Microsoft on one side and Intel on the other show a distinct divergence in PC architecture is aboutto accelerate. Microsoft has historically worked with Intel to determine the course of PC improvements and particularly the splitting up of functionality between the various parts of the system - which functions are best handled in the hardware architecture and which ones can best be handled by layers added to the operating system (OS). Microsoft has, with good cause, championed putting more of the "glue" that holds different systems together by putting "hardware abstraction layers" into the OS. In the past this has been seen in such things as the drivers that allow various peripherals to interact with the OS in a "uniform and seemless way" (yea right, "Plug and Pray"). With graphics and multimedia bursting from a nitch to mainstream part of the system architecture, many incompatibilities resulted. For instance, some games worked well only on Intel Pentium processors. Partly to remedy this and also to speed performance and evolution of the PC platform, Microsoft developed DirectX.
IDTI, AMD and Cyrix have all announced that they will not only support MMX-II in their upcoming generation of processors, but that they will also include a new set of instructions that will greatly speed 3-D graphics rendering performance beyond the capabilities of Intel's Pentium II with MMX-II. Each manufacturer will have unique instructions to speed video and audio performance and they will all be certified by Microsoft to work with Windows software. DirectX will be able to tell which MPU is in the PC and will take advanatage of its unique instruction set to deliver enhanced performance compared to Intel systems. This will increase the competition and diversity of systems available for deskstop, laptop, palm-top, NC, NA and embeded applications. I expect it to push the envelope of performance beyond competitively priced systems from Intel - further depeening the gap between the cost and performance structure between the two camps (Intel slot 1 and non-Intel). This should spell the "begining of the end" to Intel's dominance of the desktop marketplace. This is a significant development that will become apparent by the 3rd quarter of next year. IDTI looks to be in this race in a much more timely fashion than their belated entry into the Pentium 200 MMX class race. They may even have a jump on the competition - getting from the drawing boards into ramped production remains to be seen.
Perham siad that IDT is commited to enter the competition for the next generation of systems and that he has confidence in the team Glen Henry has assembled to do the job. If this groug is able to deliver, we investors will owe them our thanks because the rewards will be substantial. |
