EE Times has an article about WINHec in which the cludgy, antiquated, brain dead ISA/PCI legacy PC architecture is ready for change. MS, Intel, Compaq and others in fact could be slighted for just riding the wave in the manner of an out of control drunken free-for-all rather than a more organized, methodical progression. Having multiple sources of cheap chips, connectors and IRds to implement IEEE 1394 ("fire wire") and USB serial busses also helps. Frankly, this has been a major problem for several years that has developed because of the loose cooperation and planning between hardware and software disciplines.
These are apparently just talks at this point with the objective to develop a new PC architecture spec by this time next year. IMO, this is something that Bill Gates and fellow softies should push onto the industry with great vigor and determination. No place for mellow laid back attitudes that will mire the effort in years of debate and industry camp squabbles.
The need is there to simplify the PC experience, make it more flexible, responsive and scallable and allow for multiple, distributed processing and hardware architectures.
What this could mean to IDTI, if it evolves with input from others besides almighty Intel, is the freeing up of the systems architecture and the greater reliance on the OS to manage allocation of resources and functionality (Bill G. must be smiling about that). IEEE 1394 would serve as a primary subsystems interconnect such as between the MB and the disk drive or CD. Peripherals such as printers, plotters, keyboards and monitors would daisey-chain off of the USB port. The ISA bus would disappear and PCI would be relegated to support legacy boards and would likely disappear over time. Some systems would likely come out without PCI bus support as new peripherals came to support USB and IEEE 1394.
The firewire interface combined with smartly designed OS software (from MS? I guess if they throw 200 programmers at it they might get around to delivering it by the year 2010) has the potential to make the proprietary hardware interconnect schemes championed by monopoly minded Intel as a way of maintaining domination of the industry a thing of the past.
IBM, IDTI, Cyrix/NSM, AMD, Compaq and others who want open standards need to get behind this effort fully and not let Intel dominate them. |