Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!rutgers!sri-spam!mordor!lll-lcc!ames!oliveb!intelca!mipos3!cpocd2!howard
From: howard@cpocd2.UUCP (Howard A. Landman)
Newsgroups: comp.arch
Subject: Re: Sort Co-Processors
Message-ID: <916@cpocd2.UUCP>
Date: Fri, 16-Oct-87 14:14:17 EDT
Article-I.D.: cpocd2.916
Posted: Fri Oct 16 14:14:17 1987
Date-Received: Wed, 21-Oct-87 20:29:48 EDT
References: <112@sdeggo.UUCP>
Reply-To: howard@cpocd2.UUCP (Howard A. Landman)
Organization: Intel Corp. ASIC Systems Organization, Chandler AZ
Lines: 17

In article <112@sdeggo.UUCP> dave@sdeggo.UUCP (David L. Smith) writes:
>Has anyone out there ever run across a sorting co-processor?  I was doodling
>around the other day and it seemed to me that a lot of computers spend a
>lot of time sorting things out.

The rule of thumb at Amdahl in about 1977 or so was that commercial IBM
computing installations spend 1/2 their CPU time sorting.  Yes, that was
50%.  This created a great opportunity for products like SyncSort, which
ran somewhat faster than the standard IBM sort.  A 10% speedup in sorting
was equivalent to speeding your whole machine up by about 5%, which, on
a $4M+ machine, was worth hundreds of thousands of dollars.

-- 
	Howard A. Landman
	{oliveb,hplabs}!intelca!mipos3!cpocd2!howard	<- works
	howard%cpocd2%sc.intel.com@RELAY.CS.NET		<- recently flaky
	"Unpick a ninny - recall Mecham"