Neural networks in which the interconnections between the nodes are randomly assigned are promising for the realization of neuromorphic devices in which the resource requirements for training are lower than for a fully deterministic system. Reservoir computing is a class of recurrent network for which the input and internal weights are random and fixed over time, and only the output weights are trained via a linear regression. In this work, we review the recent work on photonic reservoirs and describe our recent results on the implementation of a single node system based on multi-mode optical interconnect technology developed for high channel density and low power data transfer applications. We discuss the potential advantages of this approach for the realization of a photonic cluster of reservoirs.
The packaging of photonic devices remains a hindering challenge to the deployment of integrated photonic modules. This is never as true as for silicon photonic modules where the cost efficiency and scalability of chip fabrication in microelectronic production facilities is far ahead of current photonic packaging technology. More often than not, photonic modules are still packaged today with legacy manual processes and high-precision active alignment. Automation of these manual processes can provide gains in yield and scalability. Thus, specialized automated equipment has been developed for photonic packaging, is now commercially available, and is providing an incremental improvement in cost and scalability. However, to bring the cost and scalability of photonic packaging on par with silicon chip fabrication, we feel a more disruptive approach is required. Hence, in recent years, we have developed photonic packaging in standard, highthroughput microelectronic packaging facilities. This approach relies on the concepts already responsible for the attractiveness of silicon photonic chip fabrication: (1) moving complexity from die-level packaging processes to waferlevel planar fabrication, and (2) leveraging the scale of existing microelectronic facilities for photonic fabrication. We have demonstrated such direction with peak coupling performance of 1.3 dB from standard cleaved fiber to chip and 1.1 dB from chip to chip.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.