Path: utzoo!attcan!uunet!husc6!mailrus!uflorida!beach.cis.ufl.edu!seeger
From: seeger@beach.cis.ufl.edu (F. L. Charles Seeger III)
Newsgroups: comp.sys.ibm.pc
Subject: Re: IBM PS/2 70 vs MAC II
Summary: DMA doesn't speed up a PC
Message-ID: <18196@uflorida.cis.ufl.EDU>
Date: 14 Sep 88 14:15:09 GMT
References: <8X7VPiyS2k-0M0m1wg@andrew.cmu.edu> <12292@cisunx.UUCP> <5489@pasteur.Berkeley.EDU> <342@apexepa.UUCP> <7057@bloom-beacon.MIT.EDU>
Sender: news@uflorida.cis.ufl.EDU
Reply-To: seeger@beach.cis.ufl.edu (F. L. Charles Seeger III)
Organization: UF EE Dept
Lines: 18

In article <7057@bloom-beacon.MIT.EDU> arwall@athena.mit.edu (Chumley Wood) writes:
|In article <342@apexepa.UUCP>, gary@apexepa (Gary Wisniewski) writes:
|>
|>Walter is correct---the Mac (even Mac II) is slower at I-O.  Although the
|>SCSI interface and drives are faster, all Macs lack DMA capability.  ...
|
|Well, that's almost true.  The Mac does have the DMA chip; Apple just
|neglected to use it...

I don't think this is the case.  XTs use DMA for hard disks, but ATs don't.
Nor do '386 AT clones.  In these machines DMA is used primarily for floppies.
It was faster to use DMA on a 4.77 MHz 8088 (8 bit bus), but faster to use
the CPU on AT and better machines -- at least assuming well written code.
Therefore, the CPU handles hard disk transfers on these machines.  Apple
may have found transfers were faster when done with the CPU rather than
with the DMA chip.  I don't know enough about the Mac to comment further.

Chuck