[Cryptech Tech] goals / use cases

[Fredrik Thulin]

On Saturday, January 24, 2015 09:19:01 PM Randy Bush wrote:
...
> so, for giggles, let's say serious tls and circuit level crypto are a
> later design, and the alpha target is dnssec signing (and validation?),
> rpki, tor consensus, and, for giggles, pgp (at human rates).
> 
> folk such as rob, jakob, and linus (on family leave) should be able to
> transform the dnssec, rpki, and tor consensus use cases into estimates
> of the number of keys, rate of change, algorithms, need for speed to
> validate, etc.

Thank you Randy, I have been thinking of how we could approach the discussion 
of specifications for the Alpha board again for a couple of days. 

It seemed like a chicken and egg problem where some people has said we need 
requirements in order to specify the technical solution, but since we're 
mostly a bunch of engineers without a PHB to come up with the requirements I 
had a hard time seeing how to address this. You anchoring one of the tents 
pins in the ground works for me =).

Could we get consensus for Randy's proposal to aim for working and secure, but 
not necessarily fast ("fast" being something we would measure in megabits/s) 
for the Alpha board?

Joachim has made a proposal for an FPGA. Although $350 per FPGA sounded pricey 
even for an Alpha board to me, I think he put forward good reasons for that 
size of FPGA so I won't object.

What methodology can we use to choose MCU? I think the choice at this time is 
either for an ARM Cortex M4 (later on perhaps upgraded to pin compatible new 
M7), or Cortex A8/A9. I think both will work for this "not necessarily fast" 
target, but there would/could be a huge difference in how we do development in 
the next year or so based on this choice.

/Fredrik

PS. I have a lead for a place in Sweden that would likely be able to assemble 
boards for us once we have a design, if having them made in Asia is 
troublesome (thinking of shipping time etc.).