These are valid questions, let me try to add some color:
1) Depends what you mean by "PDK" - the key part is establishing the end-to-end process flow, incl. FEOL, BEOL, back-end, packaging. As far as components go, leverage what the foundry offers as much as possible and optimize only the parts where you differentiate (i.e., the modulators). The messaging indicates that this stage was completed some time ago. Offering an open PDK to 3rd parties should come after you've proven out the technology in terms of capabilities (function, performance) and reliability and should not gate initial production of samples.
2) Manufacturing probably 4-6 months, testing 1 month, a good chunk of which is packaging for testing, rather than testing itself. Test development runs in parallel with design and manufacturing.
3) This should really run in parallel with all of the above, and relationships with potential takers presumably exist already.
4) Assuming that the customer has resources allocated, maybe another month or two for testing on their side for PIC-level testing. Integration into a module will take longer, maybe 3-4 months. The long pole in the test is qualification, which can take 6-9 months.
5) It's less the negotiation of the contract, but more the following factors:
- How often do you have to iterate on the above steps before you get it right?
- Is your supply chain capable of producing the volumes required, with sufficient yields to meet cost targets? This needs to be front and center from the start - 200mm should be good though for this market.
One challenge I see is in integration - these PICs are likely not reflow compatible as the reflow temperature exceeds the polymer's decomposition temperature.