Groups Home | Help | Sign in * Go to Google Groups Home Web Images Groups News Froogle Local^New! more >> [ ] [Search] Advanced Groups Search Preferences sci.crypt > How fast / slow is 1000bit ECC? - View Parsed Path: archiver1.google.com!news2.google.com!news1.google.com!newsfeed.stanford.edu!syros.belnet.be!news.belnet.be!fr.usenet-edu.net!usenet-edu.net!ciril.fr!univ-lille1.fr!news.unicaen.fr!univ-angers.fr!univ-rennes1.fr!irisa.fr!news-rocq.inria.fr!not-for-mail From: har...@pauillac.inria.fr (Robert Harley) Newsgroups: sci.crypt Subject: Re: How fast / slow is 1000bit ECC? Date: 06 Dec 2002 13:31:54 +0100 Organization: ArgoTech Lines: 18 Sender: har...@pauillac.inria.fr Message-ID: References: <6f35025c.0211201117.7697a9e0@posting.google.com> NNTP-Posting-Host: pauillac.inria.fr X-Trace: news-rocq.inria.fr 1039177922 24014 128.93.11.35 (6 Dec 2002 12:32:02 GMT) X-Complaints-To: usenet@inria.fr NNTP-Posting-Date: Fri, 6 Dec 2002 12:32:02 +0000 (UTC) In-reply-to: contini@matmail.com's message of 20 Nov 2002 11:17:36 -0800 X-Newsreader: Gnus v5.1 cont...@matmail.com (Scott Contini) writes: >If we take a leap and make the assumption that we could get 1 GHz Pentium >timings for all curve sizes by multiplying by 19/5.46, then Harley's >163-bit PIII time would be 3.13 milliseconds Actually we have code optimised for the FIPS sizes on x86, and a random point multiplication over GF(2^163) takes 0.682 ms. Regards, Rob. .-. .-. / \ .-. Robert.Har...@argote.ch .-. / \ / \ / \ .-. _ .-. / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / `-' `-' \ / \ / \ \ / `-' `-' \ / `-' http://argote.ch/ `-' [dot_clear] Google Home - Terms of Use - Privacy Policy - Jobs, Press, & Help (C)2005 Google