in reply to Re^10: SaltedDigest Salt?
in thread SaltedDigest Salt?

Of course.

Let me step through it for you again:

  1. This pdf shows that FPGAs can be used to perform both DES and SHA512 (and a bunch of others) encryption algorithms.
  2. Fig 12. of that pdf shows that SHA512 runs 6 times1 faster than DES.

    That experiment used a single FPGA to achieve 91 Mb/s DES.

  3. This link shows that the current record holder for performing DES encryption is The COPACOBANA device with a rate of 2^16 Mencryptions/s.
  4. I based my projections on the COPACOBANA's DES rate. But as shown, the SHA512 algorithm can be tackled at 6 times that speed.
  5. The 2008 version of the COPACOBANA device used 128 Virtex-4 SX 35 FPGAs to achieve that.

    These devices have 512 DSP slices running at 500Mhz to produce 250 Giga Multiply-Accumulate/second (GMACS) performance.

  6. Now, 4 years later, the hardware has moved on. The latest in that range is the Virtex-7 range.

    These devices have 7 times as many DSP slices (3,600) and 20 times2 the peformance (5,335GMACS).

So, going slowly now. SHA512 can be performed 6x faster than DES, and DES has achieved a rate of 2^16Me/s on hardware with 250GMACS. The latest hardware is 20x faster.

So I make that 65536e6 * 61 * 202 = 7.86432e15 SHA512 encryptions/s. Do you concur?

And to bring the story right up to date. The latest hardware, teh RIVYERA can scale to 1796 FPGAs per rack. I'll let you do the math on what its throughput could be.

With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority".
In the absence of evidence, opinion is indistinguishable from prejudice.

The start of some sanity?

Replies are listed 'Best First'.
Re^12: SaltedDigest Salt?
by zwon (Abbot) on Feb 10, 2012 at 06:34 UTC
    Pdf from the point 1 doesn't provide enough information to make any assumptions regarding relative performance of DES and SHA512. The fact that throughput of 3DES running in unspecified mode was 6 times less than throughput of SHA512, doesn't mean anything. 4 is a complete rubbish. In my view you are totally making facts up.
[reply]
      Pdf from the point 1 doesn't provide enough information to make any assumptions regarding relative performance of DES and SHA512.

      If you cannot not read a simple graph, then this discussion is pointless.

      (Notice also the highlighted text: ... the most secure function, SHA-512, is also the fastest of four investigated hash functions. ")

      And nothing is "made up". It is all laid out for you to read for yourself. If you are incapable of doing so despite the lengths I have gone to make it possible for you to verify every single bloody step, then fine. Bury your *&*&$%$£^ head in the sand.

      With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
      Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
      "Science is about questioning the status quo. Questioning authority".
      In the absence of evidence, opinion is indistinguishable from prejudice.

      The start of some sanity?
[reply]

        Ok, let's see this graph again. Fig 13 shows that throughoutput of SHA512 was 616Mbit/s comparing to 91Mbit/s for 3DES. As we talking about breaking password, we interested in number of encryptions per second, not in throughoutput. SHA512 block size 512 bit which gives us 1203125 encryptions/s. 3DES block size 64 bit, that's 1421875 3DES encryptions/s or 4265625 DES encryptions/s. And this turns your 6 coefficient into 0.28. Now, reading description of tests in you can see that they were hashing 3MB of data, which means that they probably used 3DES in CBC mode, so couldn't use many optimisations, like e.g. parallel encryption of several blocks. Note also that Fig 12 shows that SHA512 implementation used 23% of FPGA slices, twice as much as SHA1 implementation (not counting IO slices), and document says nothing about 3DES.

        Summarising:

        1. SHA512 in [1] was 3.5 times slower than DES
        2. you're not taking into account number of slices required to implement SHA512 and DES
        3. [1] is about hashing data stream, copacobana is about key search. [1] doesn't allow parallel operations, copacobana does.

        And based on that I think your estimation sucks

        [1] http://www.east.isi.edu/~bschott/pubs/grembowski02comparative.pdf

[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom