[Cryptech Tech] Alpha board strategy

[Peter Gutmann]

Ben Laurie <benl at google.com> writes:

>> but I also don't think going with that is a very good idea.
>
>Pah. Any particular reason?

>From a business point of view, you're asking manufacturers to go with an
experimental research CPU implemented as a soft-core rather than a standard
SoC or IP core on an FPGA, using experimental patches (not in the mainstream
branch) to an existing compiler.  That kinda screams "run away" to any
commercial vendor.

>From a practical point of view, you now need to write (or modify) code to take
advantage of the capabilitie features rather than using off-the-shelf code,
and figure out why it doesn't work, and possibly debug the tools you're using
to work with it.  See above for "run away".

That's not to say that CHERI isn't a nice piece of work, but there's quite a
gap between "cool proof of concept" and "ready for commercial prime-time".

Peter.