I'm not sure about that. Sure they may do some preliminary work to assess the technology, but I would think that any proper design work would be on a separate contract (to allow them to actually USE the IP, rather than just assess it) and this would be considered a material milestone and require an announcement in order to comply with their continuous disclosure commitments. At a minimum it would be a letter of intent. I don't think they could do it in "secret", and I don't see why WBT would want that done secretly. Surely they'd shout it from the rooftops as soon as they could. The only thing I'm not sure of here is how that would apply to customers using designs in Skywater's standard design library. Perhaps Skywater is considered the customer in this sense?
This would apply for the first couple of customers as future customers would not necessarily be as significant a milestone (unless it's a massive customer).
I won't add much else to what
@writer has said, other than to say that I'm not really putting too much emphasis on the DoD potential. Yes it's cool to be associated with them, but I can't see it being big enough volume compared to a consumer/industrial product customer. I've nothing to back that up other than gut feel though.
I also think the really big fish for WBT is the discrete market (pretty sure Coby has said this as well), as it will allow them to offer their own memory drives to sell as their own product, rather than just getting royalties. While the margins may be lower on this, the sheer amount of $$ this would bring in would dwarf the royalties income. Think of them supplying SDD's to all the data centres around the world as their current flash drives reach the end of their life. The power savings are also a massive selling point here. While work is no doubt progressing as fast as they can on this, they will not be rushing it to bring an imperfect product to market before it is ready.
The other thing to consider is that in the grand scheme of things, Skywater is a very small fish in a very big pond (in terms of production capacity). While production there may start 2024, they will need to do the technology transfer to every fab they sign up. This means when the big players do eventually sign up, it'll still be 12-18months before they are producing from when they sign up. I'm not sure how much time they can cut out of this due to the learning curve from doing it at Skywater, but I don't think it would be much less that the 18months it's taking now (simply because actually manufacturing the chips takes so long). The good thing in their case though is that the customers of the big players should be able to start designing straight away once they start the tech transfer, as they will have the confidence that ReRAM can be successfully mass produced as shown by Skywater.
(20min delay)