US 12,189,559 B2
Methods, apparatus, and articles of manufacture to reorder N-dimensional sparse data into groups of data elements that can be collocated in a memory
Anirud Thyagharajan, Bengaluru (IN); Prashant Laddha, Bengaluru (IN); Om Omer, Bangalore (IN); and Sreenivas Subramoney, Bangalore (IN)
Assigned to Intel Corporation, Santa Clara, CA (US)
Filed by Intel Corporation, Santa Clara, CA (US)
Filed on Jun. 26, 2020, as Appl. No. 16/913,370.
Claims priority of application No. 202041018794 (IN), filed on May 2, 2020.
Prior Publication US 2020/0327396 A1, Oct. 15, 2020
Int. Cl. G06F 13/40 (2006.01); G06V 10/82 (2022.01)
CPC G06F 13/4013 (2013.01) [G06V 10/82 (2022.01)] 25 Claims
OG exemplary drawing
 
1. An apparatus comprising:
memory;
instructions stored in the memory; and
one or more processor circuits coupled to the memory, the one or more processor circuits to utilize the instructions to:
reorder N-dimensional sparse data into a chunk of spatially collocated data, N being a positive integer, the reordering including:
identifying occupied data elements of the N-dimensional sparse data; and
identifying adjacent ones of the occupied data elements;
create a graphical representation of the occupied data elements where a node of the graphical representation represents an occupied data element and an edge that connects nodes of the graphical representation represents an adjacency between data elements represented by the nodes; and
forward the chunk for convolutional neural network processing.