Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!iuvax!purdue!mentor.cc.purdue.edu!pur-ee!hankd
From: hankd@pur-ee.UUCP (Hank Dietz)
Newsgroups: comp.arch
Subject: Re: VLIW Architecture
Summary: History?  Things are rarely so clear.
Keywords: VLIW
Message-ID: <13109@pur-ee.UUCP>
Date: 9 Oct 89 23:03:32 GMT
References: <251FCB3F.12366@maccs.dcss.mcmaster.ca> <1050@m3.mfci.UUCP> <13050@pur-ee.UUCP> <1630@l.cc.purdue.edu> <1989Oct5.025841.2046@esegue.segue.boston.ma.us> <3449@alliant.Alliant.COM> <1071@m3.mfci.UUCP>
Reply-To: hankd@pur-ee.UUCP (Hank Dietz)
Organization: Purdue University Engineering Computer Network
Lines: 64

In article <1071@m3.mfci.UUCP> colwell@mfci.UUCP (Robert Colwell) writes:
>In article <3449@alliant.Alliant.COM> lewitt@Alliant.COM (Martin Lewitt) writes:
>>Let's get the history right.
>>       1) The first commercially available VLIW machine was the FPS-164,
>>          not the Multiflow (by 6 years).
>
>The FPS machines do indeed enjoy a unique place in the computer hall of
>fame, but the label under the exhibit won't say VLIW, nor should it.

I quite agree that the FPS-164 isn't trace-scheduled etc. as a VLIW, but I'm
not sure that means the hardware isn't VLIW....  However, it isn't the first
in either case.  Way back, Burroughs had some nifty WCS stuff that probably
qualifies as the first "VLIW hardware," but Multiflow's Trace clearly
deserves the title "first VLIW computer system."

>>       2) The "first affordable supercomputer" was the FPS-164, not the
>>	  Convex (by 4 years).  FPS was using the "affordable
>>	  supercomputer" and "departmental supercomputing" phrases
>>	  long before the Convex advertisements and literature took
>>	  them up.
>
>Agreed.  Although calling an attached processor a "computer" smacks of
>marketing hype, to me.

Define "supercomputer."  Intel's IPSC tried to claim the first title, and
there are plenty more if you define "affordable" the right way.  ;-)

>>       3) The first commercially available machine to compile complete
>>          HLL applications to micro-code was once again, the FPS-164

Were "complete HLL applications" really compiled to microcode?  I thought
they ran partly on the host machine?

>>       4) The first commercially available machine to successfully exploit 
>>          parallel processors automatically using "dusty deck", serial 
            ^^^^^^^^
>>          FORTRAN, the Alliant FX8, (by 4 years and counting).
>
>"And counting"?  What does that mean?

Are you defining "parallel" to be MIMD, but not SIMD/Vector?  In either
case, doesn't Cray count?  I don't know what compilers have been marketed,
but I know of more than a few "dusty deck" compilers for MIMD targets....

>>       5) The first commercially available RISC machine was the FPS-164.
>>          (I'd love to see this one discussed, are VLIWs RISCy?)   8-)
>...  The important thing is
>to be able to implement the most-used ops such that they're as fast as can be,
>which usually means they get hard-wired control.

Most early machines looked pretty RISCy...  in exactly the sense stated
above.  The main difference in recent RISC design is the use of compiled-code
statistics rather than applications algorithms in determining which ops are
frequently used.  I see no reason to single-out the relatively-modern
FPS-164; of course, just as the "first VLIW" title goes to whoever coined
the name, "first RISC" really belongs to UCB.

VLIW does tend to imply RISC...  and RISC implies VLIW if you really think
about it.  After all, parallelism is the *OBVIOUS* way to speed-up things,
and the compiler emphasis of RISC + parallelism = VLIW.  Not that the full
VLIW concept is trivial, rather that it is natural & elegant.  Best of all,
it even seems to work pretty well.  ;-)

						-hankd@ecn.purdue.edu