From owner-freebsd-hackers Mon Oct 27 01:47:45 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id BAA00752 for hackers-outgoing; Mon, 27 Oct 1997 01:47:45 -0800 (PST) (envelope-from owner-freebsd-hackers) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id BAA00735 for ; Mon, 27 Oct 1997 01:47:41 -0800 (PST) (envelope-from root@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id BAA24643; Mon, 27 Oct 1997 01:48:47 -0800 (PST) Message-Id: <199710270948.BAA24643@implode.root.com> To: sthaug@nethelp.no cc: jamil@trojanhorse.ml.org, hackers@FreeBSD.ORG Subject: Re: Parity Ram In-reply-to: Your message of "Mon, 27 Oct 1997 08:24:14 +0100." <2722.877937054@verdi.nethelp.no> From: David Greenman Reply-To: dg@root.com Date: Mon, 27 Oct 1997 01:48:47 -0800 Sender: owner-freebsd-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >ECC based on parity memory is fairly inexpensive. You need 12.5% more >bits, total, and a chipset that can handle ECC. This lets you correct >all single bit errors and detect all double bit errors (also many >multiple bit errors). Actually, it's not a percentage. It works out like this: data_bits ECC_bits 8 5 16 6 32 7 64 8 As you can see, at 64 data bits the number of extra bits needed to implement ECC is the same as is needed for byte parity. -DG David Greenman Core-team/Principal Architect, The FreeBSD Project