[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: algorithms processing time
On Thu, 24 Sep 1998 dbastien@galea.com wrote:
> > We are interested in the processing time needed for encoding/decoding
> > data conforming to the DES, IDEA, RSA, MD5 algorithms.
>
> [Schneier] B. Schneier, "Applied Cryptography Second Edition",
> John Wiley & Sons, New York, NY, 1995. ISBN 0-471-
> 12845-7
This book has _very_ outdated performance figures.
> [Schneier97]B. Scheier, "Speed Comparisons of Block Ciphers on a
> Pentium." February 1997,
> http://www.counterpane.com/speed.html
While mostly the data in this paper is (almost) correct, they base on
estimations. In particular, IDEA's performance is strongly overestimated.
For the symmetric cryptosystems (incl hash functions), the
best available performance data for Pentium is available from
Bart Preneel and Vincent Rijmen and Antoon Bosselaers, "Recent
Developments in the Design of Conventional Algorithms", Computer Security
and Industrial Cryptography, State of the Art and Evolution, LNCS 1528 (to
appear). The figures are also available from
http://www.esat.kuleuven.ac.be/~bosselae/fast.html
On Pentium 2 or even on Pentium MMX, the speed of some cryptosystems may
boost. See for example my paper "IDEA: A Cipher for Multimedia
Architectures?" in the proceedings of SAC '98, LNCS (to appear). The speed
figures are available from http://home.cyber.ee/helger/fastidea/.
And last but not least, Eric Young's libeay has P2 optimized code for
several ciphers.
Helger
References: