AES256 vs Twofish performance (Was: twofish keysize)

[Malte Gell]

Am Donnerstag, 22. April 2004 09:01 schrieb Per Tunedal Casual:

> I was more concerned about encryption. What about the performance
> AES-256 compared to TWOFISH for very large files with GPG? AES makes
> more rounds for larger keys, TWOFISH does the same number of rounds
> for all key lengths. That might make TWOFISH attractive for large
> files.

AES-256 has 14, Twofish(256) has 16 rounds. BTW, there's some nice 
Information about AES at Wikipedia
http://en.wikipedia.org/wiki/Advanced_Encryption_Standard

Just made a quick "real-worl-test", test.tar has about 55 MB and is a 
tar'ed /home directory on a Linux box with different types of data. 
Actually Twofish was really slightly faster, this is interesting since 
Rijndael's believed speed was one of NISTS's major reasons why it 
became AES and 2Fish-256 has more rounds.

*AES-256*
time echo test | gpg --passphrase-fd 0 --cipher-algo aes256 --symmetric 
test.tar

real    0m28.687s
user    0m26.280s
sys     0m0.930s

*Twofish*
time echo test |gpg --passphrase-fd 0 --cipher-algo twofish --symmetric 
test.tar

real    0m27.055s
user    0m24.800s
sys     0m1.000s

It shows the difference is negligible and AES/Rijndael doesn't seem to 
be the top performer it is sometimes said to be, at least with large 
files. I think I may later make a test with a 1000 MB tar-file to see 
what happens with really large files.

HTH
Malte
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: signature
Url : /pipermail/attachments/20040422/4da58e30/attachment.bin