REGISTER | LOGIN
Breaking News
News & Analysis

OFC Debates Road to Photonic ICs

Big players bet on silicon photonics
3/23/2017 11:00 AM EDT
3 comments
NO RATINGS
< Previous Page 2 / 2
More Related Links
View Comments: Newest First | Oldest First | Threaded View
rick merritt
User Rank
Author
Re: Integration
rick merritt   3/26/2017 11:07:08 AM
NO RATINGS
@apaDAV: As I understand Smit, he is not advocating for integrating InP and silicon. Rather his is advocating using the kinds of techniques that have made CMOS successful--shared foundries with agreed-on design rules and standard PDKs and standard libraries to enable fabless InP companies.

walken1
User Rank
Author
Re: Integration
walken1   3/24/2017 12:14:49 PM
NO RATINGS
My understanding is that InP integration can be achieved through direct epi or through wafer bonding though there are challenges with the direct epi approach. This article details a wafer-level die transfer process for bonding InP laser dies onto silicon using adhesive wafer bonding. 

http://www.semiconductor-today.com/features/PDF/SemiconductorToday_JulyAugust2014-Enabling-silicon.pdf

apaDAV
User Rank
Author
Integration
apaDAV   3/23/2017 3:37:24 PM
NO RATINGS
I did not understand how InP is integrated with regular Si processes.
Does it grow on Si? Is it grown besides the logic or on top of the interconnects?
Why isn't this a packaging thing? If they can put multiple ICs in one Package, why not stack the optronics on top?
The way I understand it there is no CMOS with InP, only NMOS.

Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed