US 11,886,974 B1
Neural network chip for ear-worn device
Andrew Casper, Eau Claire, WI (US); Matthew de Jonge, Brooklyn, NY (US); Igor Lovchinsky, New York, NY (US); Nicholas Morris, Brooklyn, NY (US); and Jonathan Macoskey, Pittsburgh, PA (US)
Assigned to Chromatic Inc., New York, NY (US)
Filed by Chromatic Inc., New York, NY (US)
Filed on Aug. 11, 2023, as Appl. No. 18/232,854.
Claims priority of provisional application 63/514,641, filed on Jul. 20, 2023.
Int. Cl. G06F 7/544 (2006.01); G06N 3/044 (2023.01); H04R 25/00 (2006.01); G06F 9/50 (2006.01)
CPC G06N 3/044 (2023.01) [G06F 7/5443 (2013.01); G06F 9/5016 (2013.01); H04R 25/507 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A hearing aid, comprising:
a neural network chip, comprising:
a plurality of tiles arranged in a tile array, each tile comprising memory, a number of multiplier-accumulator circuits (MACS) in a range of 16-128 MACs, and routing circuitry;
wherein:
the memory of each tile is configured to store a portion of elements of a matrix A comprising weights of a recurrent neural network;
each tile is configured to receive and store elements of an activation vector X, wherein the activation vector X is derived from an input audio signal;
all or a subset of the plurality of tiles are configured to perform a matrix-vector multiplication A*X by performing multiply-and-accumulate sub-operations in parallel among all or the subset of the plurality of tiles;
routing circuitries of tiles in each respective row of tiles in the tile array are configured to combine results of the multiply-and-accumulate sub-operations; and
all tiles in a column of the tile array are configured to receive same elements of the activation vector X.