Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site watcgl.UUCP
Path: utzoo!watmath!watcgl!jchapman
From: jchapman@watcgl.UUCP (john chapman)
Newsgroups: net.micro,net.micro.pc
Subject: Re: sanyo 555
Message-ID: <1265@watcgl.UUCP>
Date: Thu, 14-Feb-85 10:38:27 EST
Article-I.D.: watcgl.1265
Posted: Thu Feb 14 10:38:27 1985
Date-Received: Fri, 15-Feb-85 03:40:53 EST
References: <288@ssc-bee.UUCP> <448@hou2e.UUCP>
Organization: U of Waterloo, Ontario
Lines: 35
Xref: watmath net.micro:9378 net.micro.pc:3324

> <>
> 
> I have a Sanyo on loan to me and I have the following comments:
> 
> 	1) If your looking for an IBM PC compatible, forget it. Most
> 	   everything that runs on the PC will not run on the Sanyo.
> 
> 	2) The Sanyo is slowwwwwwwwww. I believe that disk access is
> 	   not DMA driven and hence a pain to wait for.
> 
> 	3) The Sanyo is cheap... so if your not looking to be in the
> 	   mainstream of PC computing, this might be the machine for
> 	   you.
> 
> 					Dave Bloom
> 					Bell Labs, Holmdel

 Regarding (2).  Why should DMA have anything to do with it?  To my
 knowledge MS-DOS does not allow IO/compute overlap - when your
 program does a disk request control does not return until the transfer
 is complete and the elapsed time for a disk transfer is not going
 to decrease because of DMA! DMA might help in the case of a hard
 disk since programmed io is not fast enough to keep up with disk
 transfer rates most non-dma controllers buffer a sector and then the
 device driver copies the buffer contents to wherever they are
 supposed to go - this does take noticably longer than DMA transfers.
 A more likely explanation for the slowness you percieve is either
 1. they supply slow drives (also single density will have a lower
    transfer rate than double density); or
 2. the basic clock speed is lower than the machines you are 
    inplicitly comparing the Sanyo to.
 
 A purely personal opinion : it is still more cost effective,
 particularily in the long run, to go with an IEEE696 (S100)
 system.