In case anyone is curious about the Advanced RISC Computing specification’s ownership, it turns out that during the initial development phase the preliminary specification is jointly owned by Compaq Computer Corp, Digital Equipment Corp and MIPS Computer Systems Inc – no surprise there. When it’s finished, no one company will own it and it can be bought for a nominal fee. If the latest round of ACE Initiative recruits look like a bunch of refugees from the personal computer arena it’s because they’re being wooed to ACE’s side by its siren song of easy migration from Intel Corp architecture to MIPS RISC, the warm and comfy feeling they get from being able to tell users their current Intel investments won’t be wasted and ACE’s apparent initial focus on the desktop. However, the Initiative has some up-market intentions as well. An ACE Multiprocessing Special Interest Group has already been formed, led by Ing C Olivetti & Co SpA out of its research and development centre in Santa Clara, California. It is to submit its work to the main group for review in the next few months. Besides that, MIPS vice-president Jim Billmaier says the company expects to be able to produce a 1,000 MIPS fault-tolerant machine next year stringing four to 32 processors together. The MIPS R4000 chip, expected to support all this once it gets into production and forecast to perform initially at around 50 MIPS, could be stretched to do 200 MIPS.

to be its Achilles heel

Publicly, the ACE people only touched in the briefest possible way on ARC’s novel Hardware Abstracton Layer. Privately, however, one of their number called it a noble experiment, admitting that it could prove to be its Achilles heel. Intended as a means for the ACEs to differentiate themselves and add value, it’s a great concept, he said. But what happens if some of them – me too companies to their very core – lapse into chronic clone-ism and simply start knocking off the big boys’ boxes? On the other hand, he said, the whole Hardware Abstraction Layer notion is fraught with the possibility of creating serious non-compatibility problems for the whole Initiative.