Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!think!mit-eddie!genrad!decvax!ittatc!ittvax!jeffd
From: jeffd@ittvax.ATC.ITT.UUCP (Jeff Denenberg)
Newsgroups: net.micro,net.micro.pc
Subject: Re: Increasing the speed of IBM-XT.
Message-ID: <597@ittvax.ATC.ITT.UUCP>
Date: Fri, 30-May-86 09:26:03 EDT
Article-I.D.: ittvax.597
Posted: Fri May 30 09:26:03 1986
Date-Received: Sun, 1-Jun-86 07:45:30 EDT
References: <6685@utzoo.UUCP> <1300@eagle.ukc.ac.uk> <515@ecn-pc.UUCP>
Organization: ITT-ATC, Shelton Ct.
Lines: 30
Xref: linus net.micro:13367 net.micro.pc:8071

> In article <1300@eagle.ukc.ac.uk> rde@ukc.ukc.ac.uk (R.D.Eager) writes:
> >In article <6685@utzoo.UUCP> jackson@utzoo.UUCP writes:
> >
> >The V20 is usually a plug in replacement and very cost effective. I have
> >one in a real PC, and a V30  (8086  relative)  in  a  PC  compatible. No
> >trouble,  perceived  improvement  of  around  25%;  cost  around  $15. A
> >bargain. The occasional system has trouble with disk formatting  but  at
> >that price it's worth trying.
> 
>     Your perceptions may be faulty - I have measured the difference at 
>     5 to 10% on what are for me fairly representative programs.
>     Still, that is not too bad for $15 and five minutes work.

     My experience with the V20 has also been good. Its improvement is better
than a benchmark of 5 to 10% would indicate. Many critical (but usually
unmeasured) operations are sped up by a factor of two. These include:

     a memory test done at system cold boot time,

     my THESYS FASTCARD IV (2Mbit expanded ram, ramdisk, disk cache,
        printer buffer)

and 

     interactive screen operations (an important human factors issue).

     The only problem I saw was a hardware test done at boot time on my clone
(an ITT Xtra). It is speed sensitive, fails intermittently, and requires
a patch. All things considered, I suggest all 8088 or 8086 based pc clones
would benefit from this upgrade.