Re: MD5 performance

[Ran Atkinson <atkinson@sundance.itd.nrl.navy.mil>]

MD5 Performance update:

	On the 64-bit RISC processor that got 20 Mbps using the
RFC-1321 code, there is new news that by hand tweaking the code in a
one can get about triple the speed.  Simply swapping the RFC-1321 code
out and substituting the Karn MD5 code reportedly about doubles the
speed.  I'm being deliberately vague here on exact performance because
slightly different system configurations were used by different groups
to get these numbers.

  My concerns over MD5 performance are significantly abated by these
more recent results.  It looks like IP/FDDI, IP/HyperChannel, IP/ATM,
and IP/PPP/SONET may have performance problems with MD5.  Most users
are at Ethernet speed or below and will be fine.  IMHO, we need to
consider selectively authenticating packets in any event -- protect
"important" packets for some metric of "important" (perhaps source
routed packets for example).

Ran
atkinson@itd.nrl.navy.mil

---

Follow-Ups:

• Re: MD5 performance
• From: Masataka Ohta <mohta@necom830.cc.titech.ac.jp>

References:

• Re: Changes in AP draft
• From: Hilarie Orman <ho@cs.arizona.edu>

---

• Prev by Date: Re: Changes in AP draft
• Next by Date: Re: MD5 performance
• Prev by thread: Re: Changes in AP draft
• Next by thread: Re: MD5 performance
• Index(es):
  • Main
  • Thread