Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!princeton!caip!topaz!harvard!ut-sally!im4u!oakhill!davet
From: davet@oakhill.UUCP (Dave Trissel)
Newsgroups: net.arch,net.micro.att
Subject: Re: AT&T MIPS claim (Task switch instruction)
Message-ID: <709@oakhill.UUCP>
Date: Fri, 20-Jun-86 02:10:37 EDT
Article-I.D.: oakhill.709
Posted: Fri Jun 20 02:10:37 1986
Date-Received: Sun, 22-Jun-86 02:58:18 EDT
References: <577@scirtp.UUCP> <124@bakerst.UUCP> <583@scirtp.UUCP> <585@scirtp.UUCP> <206@njitcccc.UUCP> <6826@utzoo.UUCP>
Reply-To: davet@oakhill.UUCP (Dave Trissel)
Organization: Motorola Inc. Austin, Tx
Lines: 25
Xref: watmath net.arch:3527 net.micro.att:1315

In article <6826@utzoo.UUCP> henry@utzoo.UUCP (Henry Spencer) writes:

>> ... the 780 is
>> an all purpose performance machine (for its day).  It was designed
>> to handle multitasking multiprocess environments well.  It has
>> a one instruction "switch task", which I believe the 68000 lacks...
>
>The 780 is also notorious for the interesting property that many of its
>complex instructions are actually *slower* than the equivalent combination
>of simple 780 instructions.  I'm not sure about "switch task", but it would
>not surprise me if the 68000 does it faster.  ...

I don't know about the VAX 780 instruction times, but an identical situation
currently exists in regard to the Intel 386 and the Motorola MC68020.  Intel
in advertizing and seminars brags about their task switch instruction
capability and gives a time of 17 microseconds with the implication that this
obviously beats chips which don't have such a capability.  What is quite
fascinating is that I coded up the exact equivalent code for the MC68020 and
MC68851 combination and it came out to between 12 and 13 microseconds.  (Both
chips at 16 Megahertz and no wait-state memory.)

The MC68020/851 code was about 12 instructions yet definitely faster.

  --  Dave Trissel  Motorola Semiconductor, Austin, Texas
	{seismo,ihnp4}!ut-sally!im4u!oakhill!davet