Quan Gork, I think Suresh refers to the VCSEL transceiver as such, irrespective of any particular timelines or details. Whatever was needed in term of innovation, has been invented until April, with the mass-producible resonant cavity detector as its tentative culmination.

Everything needed since then was and is is engineering. It may be hard at times, but it can be done. At least that's my understanding.

That said, let me quote the Q2, Q3, and Q4 targets. I copied them from the POET materials into the POET Technologies Timeline so they can be found easily. And let's remember we are a few weeks delayed.

• Q2: "Functional transmit and receive components. – Demonstrate integrate transmit and receive functions, integrated flow with FETs, VCSELs and detectors." – My understanding: This is a proof of concept showing that the integrated stuff works in principle. Now optimization and fine-tuning are required to turn this into a product (or two products, see below) for manufactururing and selling.
  
• Q3: "Tape out first transceiver prototype – First pass design for 10 Gbps and 25 Gbps transceivers" – My understanding: Tape-out means you have everything ready to go into production. In principle you could pass a pile of files (a "magnetic tape") with all the materials to the manufacturer's foundry (here: Wavetek), you could say "build this for me", and walk away. By the end of Q3 POET wants to have the tape-out for two products completed: a transceiver for 10 Gbps and a second transceiver for 25 Gbps.
• Q4: "First integrated transceiver prototypes. – First demonstration of integrated transceivers at 10 Gbps, enables start of customer validation." – My understanding: Preparations and production in the foundry take some time (and POET probably won't walk away). By the end of Q4 they want to hold the first batch of the 10 Gbps integrated VCSEL transceiver chips in hand and give/sell them to customers for evaluation. And don't be misled by the word "prototypes"! Subhash explained that he uses that word to mean specimen of the real thing, not for a proof of concept.