Anon (no.delete@this.spam.com) on July 23, 2020 10:25 am wrote:
> Michael S (already5chosen.delete@this.yahoo.com) on July 23, 2020 12:46 am wrote:
> > Actually, looking at Intel's previous non-x86 architecture of the decade that preceded a start
> > of Itanium project, one could expect something pretty similar to Itanium, just with fixed 32-bit
> > instructions width instead of 3-verbs-per-128 and without pervasive predication.
> > Intel's non-x86 architects obviously liked Berkeley-style RISC (see i960) and obviously liked explicit
> > parallelism on the FP side of machine (see exposed-pipeline mode of operation of i860).
> > Quite possibly, Intel's own IA-64 would be even binary compatible with application-level i960 code.
>
> Didn't Intel stop marketing i960 due to an agrement with DEC?
> Wouldn't those legal questions impact any non-x86 lines?
>

I don't know.
My impression was that death of i960 was, indeed, a result of agreement with DEC, but indirect result of it, rather than direct. Sort of reverse-NIH syndrome. Anything that came from DEC, like StrongArm, was considered by upper management superior to any internally developed technology, like i960. Despite the later doing quite successfully in the market and the former having approximately zero market.
But quite possibly I overlooked legal issues.