The statement I referenced was in the R&R report dated Aug 7, 2017 which points to expanding the available passive discrete component products including the MUX/DEMUX functionality.

Let’s try to put the whole statement that I quoted yesterday in perspective re the early next year (2018) with my comment “that development progress appears to be very much in line with commercial qualification results being reported”.

POET released information about the optical interposer which maybe we can think of as the wiring harness that connects both optical and electrical circuits to and from the active and passive elements of the optical engine in a highly efficient way using a high volume microfabrication process.

Now let’s quote the news release:

Poet also announced that it has successfully completed development of the company's low-loss dielectric waveguide stack and is readying the technology for transfer to manufacturing. As part of its continued technology development, Poet has demonstrated a 10-times improvement in the optical loss of its waveguide stack as compared with the performance publicly disclosed at the Photonics Integrated Circuit (PIC) conference in March, 2017. Optical transmission loss within the dielectric stack has now been measured to be below 0.2 decibel per centimetre, which renders the stack virtually loss-less for most practical applications. The ability to deposit low-loss and low-stress dielectric waveguides is the first critical step to enabling complex functionality of these waveguides in key applications, such as arrayed waveguide gratings (AWGs), Echelle gratings, Laterally Coupled gratings, multimode interference couplers and splitters, and pass-through waveguides. These functional features are integral to enabling the performance of a transceiver optical engine and essential to the performance of the multiplexer/de-multiplexer component required in wavelength division multiplexing (WDM) systems.

If we combine this information with the news release issued the following day it points to the integration of the PIN diode with commercial metric qualification for the 25G associated with the 100G ROSA which they confirm is on target as well as the ability of the photodiode to support data rates required for 200G and 400G. In addition it also highlights operation within the two important wavelength windows which allow for  transmission applications within the data centre and outside of the data centre.

There has never been a question about Denselight’s ability to produce well designed components the key has always been to build well designed components for high demand long life cycle products.

It is interesting to speculate how POET is proceeding to enable high volume production in facilities not under their control. One approach that I understand is used by some companies is to install their own equipment (at external fabs) for critical steps in the process which I believe is a method for ring fencing the technology.

We have much to learn about what they are doing but the most important takeaway from last week’s news releases is that they are advancing the technology at a good pace (equivalent to normal development) in such a way as to enable a highly efficient process to produce optical engines with a very small footprint.