On 23:05:06 Apr 01, Christian Weisgerber wrote: > > It mentions AES but not blowfish.
Which means you are not losing anything. > There isn't any crypto accelerator that implements Blowfish. > Perhaps with reason. > Blowfish's interest is limited nowadays. It was attractive back > when the top mainstreamer cipher was 3DES, which is slow in software, Triple DES is still ridiculously slow. And AES ridiculously fast. Reason being that the mathematical beauty of AES (Rijndael) is something extraordinary. It couldn't get any simpler. Moreover one of the criteria of NIST was the ability to implement AES in hardware with simple LFSR circuits and other PLD. > but AES is just as fast as Blowfish--in fact AES was faster for > some scp(1) tests I did, if I remember correctly--and has received > more scrutiny by the cryptographic community. I have *heard* that blowfish key setup times can be quite high. Moreover I am sure that blowfish has not received the attention of cryptanalysts like AES. For all practical purposes AES is the best choice. If you are interested in doing further testing, you can enable the OpenSSL engine interface and do speed tests. IIRC it is $ openssl speed or some such thing. Best, Girish
