Energy Efficient Computing with Chip-Based Photonics (E2CDA)

The E2CDA project is a collaborative effort between Dr. Alex Gaeta (Applied Physics and Math), Dr. Michal Lipson (Electrical Engineering), Dr. Luca Carloni (Computer Science), and Dr. Keren Bergman to develop next generation application-specific optical computing systems. The student working on and leading the E2CDA project from LRL is Natalie Janosik.

With Moore’s Law slowing down, it is critical to explore alternative computing technologies to sustain the progress of chip-multiprocessor systems. We are developing photonic interconnection networks for chip-multiprocessor computing systems by capitalizing on the high parallel processing capabilities exhibited in SiP microarchitectures to accelerate target applications using electronic-optical co-design.

Ongoing applications include an opto-electronic and integrated Ising solver and Optical Matrix Multiplication Unit (OM2U) as seen in Fig. 2.

(a) target aookucation (b) accelerator
OM2U device and control system