🚀 go-pugleaf

In article <3456@alliant.Alliant.COM> lewitt@alliant.Alliant.COM (Martin E. Lewitt) writes:
>In article <1068@m3.mfci.UUCP> cutler@mfci.UUCP (Ben Cutler) writes:
>*** some deleted ***
>>VLIWs and Trace Scheduling (TM) Compiler technology go hand-in-hand.  Ellis
>>describes why the FPS-164 doesn't qualify as a good compilation target in
>>his thesis:
>
>I guess what I find strange is that the definition of a type of computers
>and a (trademarked) compiler technology are so closely tied.

That's precisely the point.  If you want to design the best high level
language possible, you don't have one team design the control structures and
a second team design the data structures and hope the two fit together well!

By analogy, if you want a high performance computer system, you want the
fastest overall system, not a superfast cpu for which you cannot generate a
compiler (and hence fast code).  So you design all the pieces together, in
this case VLIW hardware and Trace Scheduling compilation.  And by extension,
you want a real computer, not an awkward attached processor.

FPS never saw the big picture.

>From your posting, it appears Ellis correctly understood the limitations
>of the FPS architecture and his analysis of it seems to acknowledge
>some related place for it in history.

Historical footnote: The FPS machine was an attached numerical processor.
It bit the dust when general purpose computers with as much or more power and
greater ease of use appeared on the scene for many fewer dollars.