🚀 go-pugleaf

Bob Silverman writes,
>
>In article <8840005@hpfcso.HP.COM> dgr@hpfcso.HP.COM (Dave Roberts) writes:
> 
>stuff deleted..
>
>:	(6) If you really need the blazing integer speed, buy a coprocessor.
>:	    That is also one of the fundemental RISC ideas.  There are times
>:	    when things just aren't done well by software and do need hardware
>:	    help.  This option also allows you to get *really, really* fast
> 
>I'd love to. The only problem is: How do I integrate (say) an integer DSP
>chip into my SUN so it will act as a co-processor. Who modifies the compilers
>to use it, etc. etc... I could rewrite the compilers but I don't have the
>time. Do you know of a commercially available integer coprocessor that I 
>can plug into a workstation? I don't. 
>
>:	    integer speed by using a multiplier array (works by generating
>:	    all the product terms all at once and then adding the whole
> 
>etc.
>
>-- 
>Bob Silverman
>#include <std.disclaimer>
>Internet: bs@linus.mitre.org; UUCP: {decvax,philabs}!linus!bs
>Mitre Corporation, Bedford, MA 01730
>----------

Sorry Bob,
	I don't have all the answers.  I guess you do.  I was just trying
to propose some solutions.  I guess you're perfectly right.  The SPARC
is a joke.  Any machine that doesn't have integer multiply and divide is
crippled for life.  Whoops, sorry, any machine that doesn't meet your
performance requirements in any way is crippled for life, and all the
people who designed it are sick in the head.  You're right Bob.  How
could I have been so stupid.  I guess all those people who are building
chips without those functions you seem to need require are just silly,
never mind that they meet the needs of thousands of people.  Honestly,
what was I thinkin', huh?

I just thought that since you seem to have bought the wrong machine (in
your opinion), I'd try and help you make it work.  I'm sorry I even suggested
you doing any extra work.  I guess I should have included another option:
	(7) Sell the damn thing since you seem to hate it so much and buy
	    something else.  Of course this time, try running some of your
	    code on it first so you'll know that it performs up to your
	    standards.


Sorry people.  No smiley faces on this one.  I'm just too tired of trying
to deal with a bunch of whiners who'll never be satisfied with anything
I suggest.

People, RISC is not the cat's meow to everybody.  If you're one of those
for whom it doesn't work, feel free to try something else, but don't
tell me that it doesn't work for me either.

- Dave Roberts
"Just another `brain dead' designer of RISC machines"