[Cryptography] We need a new encryption algorithm competition.

Natanael natanael.l at gmail.com
Sun Mar 16 16:48:45 EDT 2014


2014-03-16 20:45 GMT+01:00 Krisztián Pintér <pinterkr at gmail.com>:
>
> Jerry Leichter (at Sunday, March 16, 2014, 8:23:29 PM):
>> An ever-growing percentage of fielded machines will have
>> hardware support for AES, making reaching even approximate parity
>> using a pure software implementation of some other algorithm extremely difficult to achieve.
>
> this is the best argument *against* putting direct algo support in
> CPUs. i deem the AES-NI instruction set rather harmful for the
> industry. it would be much better if we put general purpose
> instructions that help crypto. like huge register space (in the 8000
> bit range), versatile parallelism, support for GF field operations
> (prime and binary), better support for big num arithmetic, on-chip key
> storage, etc. just like GPU designers sit down with game developers
> and survey what they want, CPU developers should sit down with
> cryptographers.

Almost sounds like you want an FPGA in the CPU. So do I. :)

It would most certainly be useful for all kinds of tasks, in
particular cryptography and decoding of newer video codecs that don't
hardware acceleration (would make it much easier to speed up VP9 and
Daala, taking away the grip of h264 - and this translated to crypto
too, the grip of AES would be reduced if anything could run just as
fast). The most problematic part right now with simply throwing in an
FPGA is that they can only be reprogrammed as a whole and take a
little while to reprogram, which makes it hard to switch between tasks
or even multitask such as decrypting and decoding a live HD video
stream. Would be great to have a partially reprogrammable FPGA.


More information about the cryptography mailing list