US 12,067,492 B2
Processing for multiple input data sets in a multi-layer neural network
Dana Michelle Vantrease, Austin, TX (US); Ron Diamant, Albany, CA (US); Thomas A. Volpe, Austin, TX (US); and Randy Huang, Morgan Hill, CA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on May 5, 2023, as Appl. No. 18/144,129.
Application 18/144,129 is a division of application No. 17/951,084, filed on Sep. 22, 2022, granted, now 11,797,853.
Application 17/951,084 is a division of application No. 15/933,201, filed on Mar. 22, 2018, granted, now 11,475,306, issued on Oct. 18, 2022.
Prior Publication US 2023/0351186 A1, Nov. 2, 2023
Int. Cl. G06F 3/06 (2006.01); G06N 3/045 (2023.01); G06N 3/082 (2023.01)
CPC G06N 3/082 (2013.01) [G06F 3/0604 (2013.01); G06F 3/0644 (2013.01); G06F 3/0673 (2013.01); G06N 3/045 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A method of processing a plurality of data sets in a multi-layer neural network, the method comprising:
storing first data sets in a memory;
processing, using a first neural network layer of the multi-layer neural network, the first data sets from the memory to generate a first layer output of the first data sets, wherein processing the first data sets includes performing computations on the first data sets using an array of processing elements arranged in columns and rows;
storing the first layer output of the first data sets in the memory;
removing the first data sets from the memory;
adding second data sets to the first layer output of the first data sets stored in the memory;
processing, using the first neural network layer, the second data sets from the memory to generate a first layer output of the second data sets, wherein processing the second data sets includes performing computations on the second data sets using the array of processing elements; and
adding the first layer output of the second data sets to the first layer output of the first data sets stored in the memory.