Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!cs.utexas.edu!ut-emx!juniper!olorin
From: olorin@juniper.uucp (David Weinstein)
Newsgroups: comp.lang.forth
Subject: Re: Thinking FORTH
Message-ID: <3385@juniper.uucp>
Date: 31 Jul 88 15:51:39 GMT
References: <8807251832.AA03400@jade.berkeley.edu> <1534@crete.cs.glasgow.ac.uk>
Reply-To: olorin@juniper.UUCP (David Weinstein)
Organization: Austin UNIX Users' Group, Austin, TX
Lines: 38

In article <1534@crete.cs.glasgow.ac.uk> orr@cs.glasgow.ac.uk (Fraser Orr) writes:
>>
>>I am also fascinated by the NC4000.  4000 gates and it is effectively
>>more powerful than the 68000.  _That_ really says something for the
>>elegance and power of the FORTH philosophy.  KISS, but with the last
>>S standing for 'smart'.  I am gratified to know that the resulting
>>programs are faster than comparable traditional chips, as well as
>>smaller.
>
>It is quite interesting to hear all you forth programmers going on about
>how wonderful it is that you have microprocessors that can run your
>favourite language. I say this because chips like 68000, 80386 32000
>etc, are chips specifically designed to run languages like C.
>I would be interested to see any exprimental results comparing the
>preformance of a Forth chip running forth compared with a 68000 running
>C.
>

Pick your time and place. The Novix chip is very flawed (brain-damaged) in
fact, however even so, it is quite powerful (around 5 MIPs). However, Novix
sold the technology to Harris, who has just released the RTX chip, which
fixes *all* of the problems with the Novix, and is clocking in at 8-12 MIPs.

These chips have an assembly language (hard wired, not in uCode) which contains
many of the Forth primitives, and because of its multiple internal busses, it
can execute more than one low level Forth word in one clock cycle (including
in many cases the ability to make the return from subroutine as part of the
last primitive). 

I got a chance to play with one of these little silicon gems yesterday, and
they are fast...very very fast. What type of testing system were you proposing
to see which is faster?


-- 
Dave Weinstein
Internet: olorin@walt.cc.utexas.edu
UUCP: {ames,utah-cs,uunet}!ut-sally!ut-emx!{walt.cc.utexas.edu,juniper}!olorin