Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ames!amdcad!sun!kilowatt!raz
From: raz%kilowatt@Sun.COM (Steve -Raz- Berry)
Newsgroups: comp.sys.amiga.tech
Subject: Re: New stuff
Summary: Hard to believe Intel did it.
Keywords: 680x0's and computers in general
Message-ID: <97111@sun.Eng.Sun.COM>
Date: 1 Apr 89 01:56:39 GMT
References: <2017.AA2017@panchax> <846@savax.UUCP>
Sender: news@sun.Eng.Sun.COM
Reply-To: raz@sun.UUCP (Steve -Raz- Berry)
Followup-To: comp.sys.amiga.tech
Organization: Sun Microsystems, Mountain View
Lines: 53

In article <846@savax.UUCP> thompson@savax.UUCP (thompson mark) writes:
>In article <2017.AA2017@panchax> richard@panchax.gryphon.COM (Richard J. Sexton) writes:
>>2) Math chips
>>	From _EDN February 16, 1989, page 21_  Weitek has announced
>>	the Abacus 3168 numeric coprocessor. When teamed with a
>>	25 Mhz '030 it gives 6M Whetstones and 1.5M Flops Linpack
>>	(Double Precision). It seems to be a multiply/divide square
>>	root unit, only, so you would still need a 68881 for
>>	transcendental functions.
>Somewhat impressive, but why settle for less. The Intel 80860 (N-10) will
>absolutely blow the doors off the Weitek part and just about anything else.
> [lots of unbeliveable stuff about performance and low lovel graphics support]

>Initial release will be in the 33MHz and 40MHz speeds but 50MHz is in
>development.

Yeah, but is it fast? ;^) 

 One note, when I was at Raster we were doing a board design with the N10
to do all this fancy graphics stuff. The biggest problem we had with the
chip was that it was so heavily pipelined that if you ever had to abort
an instruction (interrupt) you would take major hits restarting the pipeline.
(We had an external cache in addition to the internal one, so you could take
a double miss...)
Add to this the fact that only static rams could keep up with the damn thing
to begin with and you got a complicated part to design around. In other words
the design would most likely have been memory or I/O bound. But if the
world were perfect and staic ram densities were equal to Drams, we would
have had a board that would have set the performance standard for the
industry. How fast can _you_ rotate a 3d space shuttle?

> The parts are currently sampling for $750. I have been toying
>with the idea of building an Amiga graphics board with one of these on it
>but the software support required is well beyond my capabilities.

Would be neet. I'd love the chance to do a high-perf dedicated graphics
board too. I'd also like to run Unix and Intuition on it too... Just
imagine the speed of Boing! or FA/18!

>Looks like Intel has finally designed something that is worth the silicon
>its on.

Hard to believe isn't it? Must have been a fluke.

>|      Mark Thompson                                                     |
>|      decvax!savax!thompson       Designing high performance graphics   |
>|      (603)885-9583               silicon today for a better tomorrow.  |

So butthead, is your mailer fixed yet?

Steve -Raz- Berry      Disclaimer: I didn't do nutin!
UUCP: sun!kilowatt!raz                    ARPA: raz%kilowatt.EBay@sun.com
"Fate, it protects little children, old women, and ships named Enterprize"