[Cryptech Tech] arm

[Randy Bush]

ok, i guess more perspective is needed for those who were not at this
week's meeting.  i am probably the worst person to try to explain this,
but here you go.

  o current verilog runs on both xilinx and altera.  there is a slight
    bias toward altera because the development environment sucks less.

  o the development platforms have been the terasic dev board, an even
    smaller ds0, and now the novena.  

  o we figure that a dozen or two folk will try the novena version in a
    few months.  but it's really a dev environment not for any real use.
    having an entire laptop inside the security boundary is a major
    tinfoil violation.

  o the intent of https://wiki.cryptech.is/wiki/Hardware is not a
    product, but rather a small single board alpha prototype that could
    actually be tried in an application such as dnssec or rpki signer,
    etc.

  o we figure maybe O(100) of these alphas might be experimented with by
    folk, so we want it to be manufacturable.  hence we look toward
    asia.

  o we wear a lot of tinfoil.  this is inside the security boundary,
    so we are very wary of extra blobs, code, ...  also, due to the
    radically open nature of the cryptech effort, it can not have code
    with any ipr viruses, gpl3, proprietary, ...

  o none of the core team felt confident they could design the board in
    a reasonable time, so we are looking for new help.  bunnie commented
    that board houses tended to specialize in arm cores, so we are
    trying to understand.

  o joachim has had positive experience with the coretex and mbed, so of
    course we are looking at them seriously.

randy