Xref: utzoo comp.sys.apple:14944 comp.os.cpm:2637
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!cs.utexas.edu!uunet!microsoft!brianw
From: brianw@microsoft.UUCP (Brian Willoughby)
Newsgroups: comp.sys.apple,comp.os.cpm
Subject: Re: Ramdisk under CPM
Summary: How fast?
Keywords: apple cpm
Message-ID: <7186@microsoft.UUCP>
Date: 31 Jul 89 19:28:01 GMT
References: <752@madnix.UUCP> <1665@sactoh0>
Reply-To: brianw@microsoft.UUCP (Brian Willoughby)
Organization: Microsoft Corp., Redmond WA
Lines: 50

In article <1665@sactoh0> ianj@sactoh0 (Ian R. Justman) writes:
>In article <752@madnix.UUCP>, rat@madnix.UUCP (David Douthitt) writes:
>> 
>> There ARE other CP/M cards besides the slowpokes from AE and Microsoft.
>> 
>> If you don't have an Applicard, perhaps its time you left 1MHz in the dust
>> and got an older card which still has yet to be beaten at 6MHz.
>> 
>> Applied Engineering, when you gonna wake up and smell the roses?  This card
>> is at LEAST 5 years older than the Z80 Plus and runs 6 times faster.
>> 
>
>Sure it beats those cards.  But what about //c owners who want to
>run CP/M?  Just tell me where one can get an Applicard-type device
>for the //c and I'll be happy knowing that you're a little bit
>rational.
>
>BTW, what about those of us who can't currently afford an
>Applicard?  What are we to do now?  Are we in the stone ages?  We
>might, but if we get the job done, and it may take a bit longer,
>who cares?!?!?  You have to realize that a lot of us can't go out

I agree with the cost vs. speed arguement, but I wonder how fast a 6 MHz
Z80 really is.  The Microsoft Softcard runs at 3.58 MHz, not 1 MHz, and
that just barely gets it up to the speed of the 1 MHz 6502.  I started
here at Microsoft after the Softcard had fizzled out, so I'm no expert on
our product.  In fact, I actually have a Taiwan peripheral clone of the
Softcard.  But I have looked into the architecture, and a Z80 takes three
cycles to transfer a byte, where the 6502 takes one.  Add to that the
restriction that the video generator uses every other half cycle at 1 MHz
to read video RAM, and then you'll see why the Softcard would see a lot of
RAM wait states when accessing main Apple memory.  When the phase 1 Apple
clock is low, the Z80 sees a clock low for 500ns, then when phase 1 is
high, the Z80 sees a few quick bursts at 3.58 MHz, but then it stops again
when phase 1 goes low again.  I think the Softcard //e is a little faster
because of the on-board 64K which is not restricted to video timing (but
is unusable as //e aux memory).  So, depending upon whether the Applicard
has its own independant memory bus (to remove the video wait states), it
might not actually be running at a full 6 MHz.  Even if it is, its not 6
times faster than the Microsoft (or AE) card - more like twice (at best).

BTW, if the Applicard comes with any documentation or schematics which
diagram its operation, I would be curious to hear how it does interface to
the Apple.  I'm always looking for the better mousetrap.

Brian Willoughby
UUCP:           ...!{tikal, sun, uunet, elwood}!microsoft!brianw
InterNet:       microsoft!brianw@uunet.UU.NET
  or:           microsoft!brianw@Sun.COM
Bitnet          brianw@microsoft.UUCP