[rc5] Question for beginner
wshull at scicom.com
Mon Oct 27 22:10:03 EST 1997
-----BEGIN PGP SIGNED MESSAGE-----
I sent this to the list earlier, but never saw it... I'm thinking
maybe I sent it from another email address and the list is too
spamproof for its own good. If you've already seen this, my
>Eric Gindrup <gindrup at okway.okstate.edu> wrote:
>> I'm sort of curious if the new clients post about 88%
>> of the old clients (in keys/sec) due the the 7/8 utility
At 07:49 PM 10/24/97 -0700, Barry Nathan wrote:
>I'm getting over 99% of the 56-bit performance with the 64-bit
>on my Mac...
Then, Joseph Fisk asked:
>"The 7/8 utility multiplier" <- what exactly is that?
I'm guessing it's supposed to take account the fact that the old keys
were 7 bytes and the new keys are 8 bytes, so the new clients would
only be 7/8 the speed of the old.
Fortunately for us, though, it *doesn't* scale like that in this
case. As long as you
stick with RC5-32/12/X, the only thing that changes for different
values of X is how long it takes to copy the key initially at the
beginning of the expansion. The expansion algorithm's loop still
does the same number of rounds (3*(2*(r+1))) = 78 in these cases.
What it does within the loop is not dependent on the key size. The
encryption/decryption algorithms just use the expanded key table, the
size of which is solely dependent on the number of rounds (r=12).
End result: it should only be very, very slightly slower than the 56-
bit-key version. <1%.
-----BEGIN PGP SIGNATURE-----
Version: PGP for Personal Privacy 5.0
-----END PGP SIGNATURE-----
Wes Shull * wes at kuoi.asui.uidaho.edu * Prince William County, Virginia, USA
Radio Amateur (KC7QCY) * Porsche 928 Fan * Ford Festiva Owner * cat lover
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.
More information about the rc5