On Wed, Dec 12, 2001 at 05:33:33PM -0600, Mark J Roberts wrote: > William Ahern: > > so why was the Tiger hash not chosen? its 192bit *and* faster than > > md5. i haven't heard anything negative about, and if somehow > > its not more secure than sha1 its still certainly better than the broken > > md5. > > I just grabbed the Tiger reference implementation and benchmarked it > against optimized SHA-1 and MD5 code: > > MD5: 53.2 MB/s > SHA-1: 26.9 MB/s > Tiger: 19.2 MB/s > > Maybe it's faster on 64-bit hardware?
hmmmm, i thought that i read somewhere it was faster than MD5. i have run my own rough tests (using Tiger and SHA-1 in a hash-based PRF) and Tiger was much faster than SHA-1, tho part of that may have been due to Tiger have a longer output. w/o evidence to the contrary, i'll concede MD5 being faster than Tiger. the designers' paper says it is 2.5 times faster than SHA-1 on 64-bit hardware: http://www.cs.technion.ac.il/~biham/Reports/Tiger/tiger/node5.html _______________________________________________ freenet-tech mailing list [EMAIL PROTECTED] http://lists.freenetproject.org/mailman/listinfo/tech
