Mo: How does this progress affect POET and our investment?

Not much, if at all. But of course competition never sleeps.

First, this is silicon photonics with all its drawbacks (→ indirect bandgap).

Second, "they believe that their beamsplitter could be ready for use in such computers within about three years, and that data centers needing faster connections may well incorporate such devices even sooner." Well, that's what the university researchers believe. Drawing from our POET experience, I believe they are underestimating the necessary monetization efforts. Well, POET Technologies is much, much closer to monetization anyway.

Third, their idea of not only moving light around inside the chip, but using light to actually process or even store data is of course intriguing. However, as of now it is nothing more than an idea. At least as as far as I know there is no device anywhere that could switch light with light. Fascinating dreams of the future! Might come some day, but not now. I would imagine the POET technology becoming the preferred platform for such developments.