Our Pitt Momentum Fund proposal entitled, “Highly Scalable and Efficient Deep Learning Accelerator Enabled by 3D Photonic Integration” has been generously funded by Pitt with additional matching financial support from the ECE department for nanofabrication (link to project page). Our lab will fabricate and demonstrate a hybrid photonic-electronic computing prototype to accelerate matrix operations used in deep learning—the current computational bottleneck in terms of speed and energy consumption. Our proposed hybrid accelerator will leverage 3D integration of a photonic integrated circuit with a CMOS image sensor to perform large-scale matrix-matrix multiplication with extreme efficiency and speed. Our approach is tolerant to fabrication variability while bypassing the requirements of high-speed electronic readout and frequent reprogramming of analog weights—three major factors that limit scalability for current analog AI accelerators.