[Cryptech Core] Hardware production?

Phil Roberts roberts at dkey.org
Mon Mar 26 21:20:19 UTC 2018


Hi,

we did a run of 25 alphas at the end of last year.  I’m looking at most of them right now.  We are waiting (seemingly forever) to get a faceplate done for the cases and then they will be ready to ship to folks other than insiders.

We are talking about doing another run of alphas real soon that will fix some minor issues and be ready for something a little bit closer to commercial production.

We have not started talking to a production house - not much of the work has been done on the changes, but we need to ramp that really soon, so now is the time to be talking about these.

Regards,
Phil



> On Mar 26, 2018, at 10:24 AM, Peter Stuge <peter at stuge.se> wrote:
> 
> Hi all,
> 
> first, many thanks for taking notes at the last f2f meeting!
> 
> 
> Then: What is the current state of hardware production runs?
> 
> I remember from the meeting before, that one production run for
> some number of unmodified Alpha boards was planned before
> end-of-year 2017, and that timing of a further production run
> was unclear, depending on what modifications were to be done.
> 
> I've done some preliminary analysis of how feasible it would be to
> create a PCI Express carrier for the Alpha, and it looks fairly good,
> but not trivial.
> 
> Mechanical: The Alpha PCB is 1.1mm too wide to meet PCI-E spec, but I
> think that violation may be OK.
> 
> Electrical: The power supply on the Alpha needs to be split up, my
> prefered solution would be to move the "intermediate" 5V regulator
> from the Alpha to the carrier, and generate low-current 16-18 V for
> the noisy diode.
> 
> Signalling: PCI-E USB host controller and two small ARMs (SAMD11 as
> mentioned before) connecting to the UART headers.
> 
> 
> So for a carrier board to work the Alpha which it carries should have
> several components not populated, which can be done at least two ways:
> 
> 1. Don't populate in PCB assembly (much preferable)
> 
> 2. Unsolder after-the-fact (risk destroying other parts)
> 
> 
> If there is a production run in the pipeline, then maybe a modified
> deliverable with certain components unpopulated could be included in
> the run? What do you think? And/or what do suppliers say?
> 
> 
> Kind regards
> 
> //Peter
> _______________________________________________
> Core mailing list
> Core at cryptech.is
> https://lists.cryptech.is/listinfo/core



More information about the Core mailing list