Support for post-quantum algorithms, roadmap?

[Mithrandir2k18]

Some post-quantum algorithms have already been published and standardized. I'm interested in buying some Nitrokeys, but if migrating to post-quantum algorithms isn't planned, I'd rather wait for a future version supporting them, as it seems like adoption is already being pushed for.

[daringer]

There is no clear roadmap as we are mostly waiting for roadmaps ;) No, seriously the main driver will of course be NIST - as also mentioned in your first link others will design their solutions based on these decisions. Furthermore Nitrokey 3 and its cryptographic operations are relying strongly on host/platform tools: gpg, opensc, fido (webauthn) and so on - not to talk about the OpenPGPCard spec. These have to go the first steps, we will follow fast - in fact we are already working on proof-of-concept implementations - but a roadmap with some reliable in-production target is currently not possible from our point of view.

[Mithrandir2k18]

Thank you, that makes sense. So since you said that the Nitrokey 3 relies on the host, is it fair to assume that once gpg etc. support them, the Nitrokey 3 (and maybe older ones as well?) will be able to store the new types of keys as well?

[daringer]

Which hardware will support which PQC mechanisms is still to be determined. Nitrokey 3 is for sure the most capable, but I would also not promise wide-support for it w/o having details which ones shall be realized on the desk. Most (if not all) PQC algorithms/mechanisms get/improve their quantum resistance through sheer size - means the keys grow massively in size therefore it is quite unrealistic that the older Nitrokeys will get PQC support.

[Mithrandir2k18]

Alright, thank you, that is a risk I'm willing to take. Closing this, since there really isn't anything for you guys to do.