he is not a beliver, well see by the end of year.
Joel Hruska • a minute ago
No, a magical heretofore unheard-of technology with supposed enormous improvements and a mesh of photonics and III-V semiconductors will not emerge triumphant and usher in a new world of prosperity and performance scaling.
That doesn't mean silicon photonics are bad or stupid. Doesn't mean III-V semiconductors or InGaAs are bad ideas for research. It means that silicon photonics, III-V, and InGaAs do not represent a solution.
According to the company website, POET technology is 10-100x faster, easily integrated into CMOS, flexible, and requires no expensive retrofit.
Do you know how I know that's BS?
Because if it actually *worked* as advertised, Apple, TSMC, Intel, Samsung, ASML, Applied Materials, GlobalFoundries, IBM, UMC, or the mainland-China conglomerate (forget their name) would have snapped the tech up *already.*
If Intel thought a company already had the secret to advancing Moore's law and returning to old scaling, they'd pay tens of billions for it. Few prices would be too high; whoever owned that tech would own the next 20-30 years of semiconductor design.