In regard to your question Poetech2 : Relative to that table, where does Poet stand, what are its metrics? 

Well POET stands right at the bottom of that table.
Importantly, this new proposal for a "Rome Architectures" by the University of Chicago and Illinois is not based on SOI (silicon on insulator), but on SiN (Silicon nitride) as BB Photonics & POET do.

 *TL (transistor lasers)

Amazing metrics read it one more time.

In this case study, we apply a new « Rome architectures » to Multi-Chip Module GPU (MCM-GPU) systems to evaluate network performance and power. Our results show that « Rome » can efficiently scale MCM-GPUs with up to 1024 streaming multiprocessors (SMs), which is >16×larger than state-of-the-art commercial GPUs (e.g., Nvidia Pascal) and 4×larger than existing research proposals. 
We also quantitatively compare « Rome » with different interconnect technologies and architectures. Our results show that, compared to state-of-the-art electrical 2D mesh, « Rome » reduces network power by up to 62% (and up to 4× with more advanced laser technologies) while improving application performance by up to 143%. Compared to architectures designed for silicon photonics, « Rome » reduces network power by >2.5× while achieving comparable application performance.

Can’t wait for our OFC products demo’s.

The optical communication industry, medias and analysts are going to have a wow moment. Oh yes !

Cheers

Link - please

Direct-Modulated Optical Networks for Interposer Systems
Mohammad Reza Jokar, Frederic T. Chong, Lunkai Zhang & Yanjing Li at Uni Chicago
John M. Dallesasse University at Uni Illinois at Urbana-Champaign

https://par.nsf.gov/servlets/purl/10184049