https://www.compoundsemiconductor.net/article/101321-ntt-team-advances-photonics-integration.html

So the magnitude of what POET has achieved recently should not be underestimated. As far as we are aware, it's the first time this has been done. Monolithic integration of all aspects of a photonic engine is huge. The industry are desperately trying to find a soultion using Silicon and this article highlights the lengths they are going to to achieve this.

The article starts with this statement. Clearly a viable option using silicon isn't on the horizon at present. 

Photonic integration is still in its infancy. One of the most serious obstacles it faces is the need to use a variety of materials to achieve different functions, many of which aren't compatible with silicon integration technology.

They highlight nanowires as a potential solution, accept they don't confine light very well. Sounds to me that this is highly complicated, time consuming and expensive. 

Addressing this problem, a group of NTT Corp. researchers in Japan came up with an approach that involves combining a sub-wavelength InAsP/InP nanowire with a photonic crystal platform, which they report this week in their paper ‘Continuous-wave operation and 10-Gb/s direct modulation on InAsP/InP sub-wavelength nanowire laser on silicone photonic crystal’ which was published in the journal APL Photonics

"A small local refractive index modulation of a photonic crystal produces strong light confinement that leads to ultrahigh-quality optical nanoresonators," said Masaya Notomi, a senior distinguished scientist for NTT Basic Research Laboratories. "We make full use of this particular feature in our work."

And the below statement highlights just how long development of new tech can take. Three years of work and still in it's infancy. Gives me hope that POET are backing the correct horse.

Back in 2014, this same group demonstrated that it was possible to strongly confine light in a sub-wavelength nanowire with a diameter of 100 nanometers by placing it on a silicon photonic crystal. At that time, "it was a preliminary demonstration of the confinement mechanism, but with our present work we've successfully demonstrated sub-wavelength nanowire device operation on a silicon platform by using this method," Notomi said.

In other words: While a sub-wavelength nanowire can't become a resonator with strong light confinement on its own, when placed on a photonic crystal it causes the refractive index modulation needed to generate the light confinement.