Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!cbatt!ihnp4!mhuxu!davec
From: davec@mhuxu.UUCP
Newsgroups: comp.sys.mac
Subject: Re: Making a Mac talk SCSI to a Sun
Message-ID: <6340@mhuxu.UUCP>
Date: Wed, 25-Feb-87 15:37:52 EST
Article-I.D.: mhuxu.6340
Posted: Wed Feb 25 15:37:52 1987
Date-Received: Sat, 28-Feb-87 05:21:17 EST
Organization: Bell Laboratories, Reading, PA
Lines: 32
Keywords: SCSI

From ephraim Wed Feb 18 08:20:59 1987
Path: wang!ephraim
From: ephraim@wang.UUCP (pri=8 Ephraim Vishniac x76659 ms 014 590)
Newsgroups: comp.periphs,comp.sys.mac
Summary: Mac SCSI is mostly software
Message-ID: <414@wang.UUCP>
Date: 18 Feb 87 13:20:59 GMT
References: <2594@phri.UUCP> <1784@hoptoad.uucp> <540211194.20192@minster.UUCP>
Organization: Wang Labs, Lowell MA
Lines: 14

In article <540211194.20192@minster.UUCP>, martin@minster.UUCP writes:
> 	The review of the Mac+ in November '86 Byte states that the
> throughput of the SCSI interface is 320K bytes per second; does anyone
> know if it is really this slow? SCSI is rated at 10M Bits (= 1.25 M Bytes)
> per second, where is the factor of (almost) four going?

It's going for CPU cycles to execute the transfers at the Mac end.
The Mac does not use DMA for SCSI, and has absolutely minimal hardware
support for the SCSI interface.  At the Mac end, each byte is transferred
to or from the SCSI bus by software.  If you're careful (i.e., fully
synchronized), this means three instructions per byte.  If you're careless
(running "blind"), it means at least one, plus some padding to tweak
your timing.

------------------------------------------------------------------------
Forwarded by:
-- 
    --->Dave Caswell
	{allegra|ihnp4|...}!mhuxu!davec
		-or-
  	davec@borax.lcs.mit.edu