Single-block collision for MD5

Re:Single-block collision for MD5 by rphlx on Sunday January 29, 2012 @04:17PM (#3526292)

Not really. This collision required specific intent, and a great deal of compute power, to find.

md5 is still adequate for accidental/random error detection because the universe doesn't (usually) spend many hundreds of hours on hundreds of GPUs to try to corrupt your data without you noticing :)

[ Parent | on HN ]

Re:Single-block collision for MD5 by snewman on Sunday January 29, 2012 @06:10PM (#3526622)

Though sometimes it can seem that way. :)

[ Parent | on HN ]

Re:Single-block collision for MD5 by 0x0 on Sunday January 29, 2012 @06:27PM (#3526668)

Fair point. I just kind of assumed that any collision would require quite a lot of flipped bits, because usually just one flipped bit is enough to cause an "avalanche" of changes in the hash sum.

Would it be possible to find a collision where only one bit is flipped in the input?

[ Parent | on HN ]