From owner-freebsd-current  Sat Jun  1 04:47:55 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id EAA27128
          for current-outgoing; Sat, 1 Jun 1996 04:47:55 -0700 (PDT)
Received: from news1.gtn.com (news1.gtn.com [192.109.159.3])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id EAA27119
          for <current@freebsd.org>; Sat, 1 Jun 1996 04:47:52 -0700 (PDT)
Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) id NAA20576 for current@freebsd.org; Sat, 1 Jun 1996 13:30:16 +0200 (MET DST)
Received: (from andreas@localhost)
          by klemm.gtn.com (8.7.5/8.7.3) id NAA00651;
          Sat, 1 Jun 1996 13:05:06 +0200 (MET DST)
Date: Sat, 1 Jun 1996 13:05:06 +0200 (MET DST)
Message-Id: <199606011105.NAA00651@klemm.gtn.com>
X-Newsreader: knews 0.9.6
From: andreas@klemm.gtn.com (Andreas Klemm)
Subject: Re: Real Time on NT/95? (fwd)
To: current@freebsd.org
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

Hi !

Does someone know, how the value is for FreeBSD-current ? Just my
personal interest.

	Andreas ///

----- Forwarded Message -----
From: Edmond <underwoe@Colorado.Edu>
Newsgroups: comp.benchmarks
Subject: Re: Real Time on NT/95?
Date: 31 May 1996 06:10:29 GMT
Organization: CNS
Lines: 25
Message-ID: <4om2gl$47p@lace.colorado.edu>
References: <31ADDDFA.1754@mntgte.lockheed.com>

Paul Maynard <paul.maynard@mntgte.lockheed.com> wrote:
>Greetings:
>
>I've been told to look into porting an application from QNX to 
>Windows NT/95.  We need to get better than 16 milliseconds 
>interrupt service response time (i.e., once a hardware interrupt 
>has been generated, we need to be in our service routine in less 
>than 16 milliseconds).  Can this be done (guaranteed) in an 
>Intel Win 95/NT environment? 16 milliseconds is a long time, and 
>in an OS like QNX its trivial.  But NT/95???? 
>
>Has anyone done a comparison of interrupt response times between 
>different OS's such as QNX, LynxOS, and NT/95 on the same Intel
>CPU?

I'm getting around 0.9 microseconds per interrupt under NT.  

-----------------------------------------------------------------------
Edmond Underwood
E-mail:  underwoe@Colorado.Edu
Bench32 1.10 final beta for Windows NT
Bench32 1.07b for Windows 95
http://www.rmii.com/~underwoe/bench32.html.