New integrated acousto-optic matrix algebra processor architecture

Abstract
A new integrated optic matrix algebra processor architecture that utilizes guided‐wave multifrequency acoustooptic (AO) Bragg diffractions is proposed. The architecture is potentially capable of performing matrix‐vector multiplications faster than any AO architecture reported heretofore, and matrix‐matrix multiplications at rates comparable to the fastest bulk‐wave AO architecture. An integrated AO processor module has been realilzed in a Y‐cut LiNbO3 substrate 1.0×10.0×28.0 mm3 in size to demonstrate the multiplication of a 4×4 matrix with a 4‐element vector at the optical wavelength of 0.6328 μm.