Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!cs.utexas.edu!csd4.milw.wisc.edu!lll-winken!ames!ames.arc.nasa.gov!lamaster
From: lamaster@ames.arc.nasa.gov (Hugh LaMaster)
Newsgroups: comp.arch
Subject: Re: What is a Mainframe?
Message-ID: <27994@ames.arc.nasa.gov>
Date: 4 Jul 89 01:39:31 GMT
References: <125@ssp1.idca.tds.philips.nl> <20752@winchester.mips.COM> <27637@lll-winken.LLNL.GOV> <164@bms-at.UUCP>
Sender: usenet@ames.arc.nasa.gov
Distribution: comp.arch
Organization: NASA - Ames Research Center
Lines: 34

In article <164@bms-at.UUCP> stuart@bms-at.UUCP (Stuart Gathman) writes:
>In article <27637@lll-winken.LLNL.GOV>, brooks@maddog.llnl.gov (Eugene Brooks) writes:
>> In article <33942@bu-cs.BU.EDU> bzs@bu-cs.BU.EDU (Barry Shein) writes:
>
>> >Mainframes are generally at the top of the line in overall I/O

While I agree that MVS is yet another incarnation of a lousy user
environment, I have to disagree with the following:

>I think that even in I/O bandwidth,  traditional  (three  letter)
>mainframes  are about to be overtaken.  A typical system has 12 x
>3 Mbyte/sec independent (and intelligent) channels.   This  gives
>an   aggregate  throughput  of  288  Mbits/sec.   The  peripheral

The Systems and Products Guide that I have in front of me shows the 3090
600S with up to 128 3/4.5 MB/sec channels.  Even the 4381 MG92E, in
IBM's "toy system" line, allows 20 channels at 3.0 MB /sec.  IBM recently
announced an option for an additional high speed channel on the 309x running at 
a measured rate of over 70 MB/sec.  (Ultra Corp. has a bus/network that 
operates at this speed.)  Total memory bandwidth available is not specified
but a quick calculation on other data listed indicates that it exceeds
465 MB/sec.  Many popular mini disk subsystems top out at less than 1-2 MB/sec
total disk throughput, and some micros do no better than 100-200 KB /sec., 
the nominal disk transfer rates of 1.8-3.0 MB/sec notwithstanding.

A second issue that has come up in this discussion is database performance.
Everyone knows that random reads and writes to disks go little faster on a
toy system than on a 3090.  So, why doesn't ACP / Sabre / VISA run on
a "PC"?

  Hugh LaMaster, m/s 233-9,  UUCP ames!lamaster
  NASA Ames Research Center  ARPA lamaster@ames.arc.nasa.gov
  Moffett Field, CA 94035     
  Phone:  (415)694-6117