It is my understanding that when Geoff Taylor worked on GaAs technology, he worked in the 980nm wavelength as there were other advantages to it. When Suresh and the rest of management came in, they were given a mandate to take POET to commercialization. During Suresh's first 90 days, he had to make an assessment and find what was under POET's hood. Having taken stock as to what was under the hood over the summer and reporting back to shareholders in October, 2015, a roadmap was developed. In that roadmap included that POET would pursue strategic acquisitions. Well that ended up as being BB Photonics and Denselight.

I believe it was announced at the THM (May, 2016) that for the next 100 days Suresh and his fellows were travelling the world to find out what customers want. Well, as we all know, that turned out to be products with 850nm wavelengths. The GaAs platform has to be recalibrated at 850nm as that is , from what I understand from the most recent AGM (July, 2017) presentation as that is what customers want. So, POET has to seek financing to recalibrate a platform to what the customers want. Now, Geoff Taylor's 980nm might be still on the table further out into the future as well, but I don't know if the industry is still going to move to that wavelength. So, opportunityknocks, it is not incumbent on current management to try to force 980nm on the industry at large. That would require a lot of investment, time, and money to turn a customer base to adopt 980nm. Why would customers risk that move now? However, customers will try GaAs at 850nm as that is compatible with current standards. POET had to consult customers for what they want "live in the field" and not what our laboratory would try, at probably great cost and effort, to woo the industry towards that standard in the near term. Maybe 980nm might have applications in different areas, other than data centres, or possibly it might be still well ahead into the future? The bottom line is that Suresh had to come up with an assessement of POET and a roadmap and one that is designed for customer needs. Well, the InP platform meets customer requirements and also does it disruptively according to manufacturing costs being lower throughout that platform.

Now if POET worked monolithically (now) in its present 980nm wavelength, the solution would work in data centres if POET devices were at both ends. In otherwords, an entire POET system. I would think that is hard to do right now as datacentres would not risk the security and performance they have from other manufactures in the 850nm space.  However, the InP platform and hybrid solutions supposedly cuts costs dramatically in that wavelength making its adoption and earlier revenue streams for POET achievable. POET knows what they are doing with the GaAs platform, it is just not what the customer wants right now. However, we have been told at the last AGM presentation that customers would like the whole GaAs solution done monolithically. That is where the whole paradigm changes and puts POET in the driver's seat or that might be why POET may be holding out for the best possible (suitable) kinds of partner(s) in the GaAs space. 

Monolithic