I would assume that once all the discrete components are qualified that just a prototypical chip will be made and qualified since every application will be different to some degree....this may shorten the qualification time, although it seems the minimum would be as lebowski suggests.

We do know that this type of tech does work on future earnings so Lazovsky's comments could be his with reference to the complete qualification of all the components and a prototype O/I developed and being tested. Specific iterations could then be developed in parallel to speed the process if confidence was high, which I believe is the case.

I think his (Lazovsky's) words should be considered carefully.