Exactly that Stitched.

Management made a strategic decision to place GaAs on the backburner, not because they considered the technology unimportant or "dead", but because the larger addressable market in the near term is with the InP products.  We are talking $billions for InP vs hundreds of $millions for GaAs, so IMO a very approriate and sound business decision when considering the order of magnitude or more  difference in the market and the company's relatively modest R&D budget.  Right now, the potential for near-term success and big money lies with the interposer/ROSA/TOSA, this is all about catching mid-wave 100G and preparing for early entry into 400G.  The GaAs will come, either via ISRO, in-house or in collaboration with someone else.