Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!necntc!ima!think!ephraim
From: ephraim@think.COM (ephraim vishniac)
Newsgroups: comp.sys.mac
Subject: Re: Accelerators beat Mac II
Message-ID: <17357@think.UUCP>
Date: 1 Mar 88 16:16:04 GMT
References: <5398@watdragon.waterloo.edu>
Sender: usenet@think.UUCP
Reply-To: ephraim@vidar.think.com.UUCP (ephraim vishniac)
Distribution: comp
Organization: Thinking Machines Corporation, Cambridge, MA
Lines: 27

In article <5398@watdragon.waterloo.edu> palarson@watdragon.waterloo.edu (Paul Larson) writes:

>Why is it that some of the 020/881 add-on cards allow the SE to
>perform faster than the Mac II?  It seems to me that a machine built
>around these two processors should be able to out-preform one which
>merely has the processors added on as an afterthought.  Is there some
>flaw in the Mac II architecture?

The Mac II is designed to allow use of 150nS memory chips.  It uses
one wait-state per memory access (if my own memory serves!).  Some of
the accelerator cards for the SE use faster memory chips and no
wait-states during memory access.  So, they're faster at the same
system clock rate.

The tradeoff is between cost and performance.  From what I read in the
trade press, I see that some of the accelerator boards use a
relatively small, very high speed cache memory to avoid the need for a
large amount of expensive RAM.  It's simpler and cheaper to build a
system with a uniform memory structure.

Whether you consider this a flaw in the Mac II architecture depends on
your thirst for speed and the depth of your pockets.

Ephraim Vishniac					  ephraim@think.com
Thinking Machines Corporation / 245 First Street / Cambridge, MA 02142-1214

     On two occasions I have been asked, "Pray, Mr. Babbage, if you put
     into the machine wrong figures, will the right answers come out?"