Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/12/84; site desint.UUCP
Path: utzoo!watmath!clyde!bonnie!akgua!sdcsvax!sdcrdcf!trwrb!desint!geoff
From: geoff@desint.UUCP (Geoff Kuenning)
Newsgroups: net.micro
Subject: Re: addresses larger than 32 bits
Message-ID: <277@desint.UUCP>
Date: Sat, 22-Dec-84 04:15:28 EST
Article-I.D.: desint.277
Posted: Sat Dec 22 04:15:28 1984
Date-Received: Mon, 24-Dec-84 02:46:47 EST
References:  <280@oakhill.UUCP> <1258@orca.UUCP>
Organization: his home computer, Manhattan Beach, CA
Lines: 20

In article <1258@orca.UUCP> andrew@orca.UUCP (Andrew Klossner) writes:

>To build a computer with 48-bit addresses and to give it 2**48 memory
>cells, you would need more cells than there are atoms in the Earth.
>Addresses bigger than 48 bits are probably not worthwhile, unless you
>need a discontiguous (sparse) address space.

I think you were thinking 10**48, Andy.  2**48 is only 256x10**12.  This is
an easily achievable number.  Terabit memories have been available for over
15 years;  you only need 256 of them to use up 48 bits.

Another way to think of this is that 2**48 is 2**32 times 2**16.  2**32 is 4
gigabytes, or about 8 Eagles.  So you need 65536*8, or 2**16x2**3, or 512K
Eagles to hold 2**48 bits.  That's probably more than the number of Eagles
that will ever be manufactured, but clearly it is a number that will be
achievable in the next decade or two.
-- 

	Geoff Kuenning
	...!ihnp4!trwrb!desint!geoff