Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uupsi!sunic!cs.umu.se!dvljrt
From: dvljrt@cs.umu.se (Joakim Rosqvist)
Newsgroups: comp.sys.m68k
Subject: Re: diff 68000 68010 (really 68010)
Message-ID: <1991Apr19.203608.4560@cs.umu.se>
Date: 19 Apr 91 20:36:08 GMT
References: <1991Apr17.114247.6446@cs.umu.se> <4900.280d4d72@miavx1.acs.muohio.edu> <1991Apr18.222550.6920@oakhill.sps.mot.com>
Sender: news@cs.umu.se (News Administrator)
Organization: Dep. of Info.Proc, Umea Univ., Sweden
Lines: 35

>In all honesty, we really do not recommend using the 68010 for new designs.
>The big advantage of the 68010 over the 68000 was its ability to handle and
>recover from virtual exceptions.  Yes it had some other minor benefits  
>(MUL & DIV timing, and 'loop mode' operation), but VM capability was by far
>the biggest.  Since the introduction of the 68020, 030, and 040 , the 68010
>offers very little value to the general computer market and has very little
>volume behind it compared to its successors.  
>
>If you are building a 'computer' we really recommend the 020 or later.  You
>can actually buy 020's today for less than 010's.  If you are building 
>an 'embedded' something and liked faster MULs, DIVs or two instruction loops, 
>our 68300 family products (68330, 68331, 68332, and 68340) all provide these
>nifty things and lots more to boot (and are also cheaper than 010s).
>
>Unlike that other members of the 68000 family, the 68010 WAS a point product 
>and we consider it to be obsolete for most (but not all) purposes.
>
>-- 
>Regards, 
>Jim Reinhart
>Motorola Microprocessor Products Group
>Austin, Texas

I was not planning to construct a new system, but instead use the 68010 as
a cheap way of speeding up an existing 68000 system (my A1000)
I would like to know *which instructions are faster* and
*by how many cycles* 
If i execute

label: move.w (a0)+,(a1)+
       dbra d0,label

on the 68k it will take 22 cycles for each loop.
What will this be with a 2 instruction cache? As good as 8 perhaps?
/$DR.HEX$