US 12,450,218 B2
Compression for sparse data structures utilizing mode search approximation
Prasoonkumar Surti, Folsom, CA (US); Abhishek R. Appu, El Dorado Hills, CA (US); Karol Szerszen, Hillsboro, OR (US); Eric Liskay, Folsom, CA (US); and Karthik Vaidyanathan, San Francisco, CA (US)
Assigned to Intel Corporation, Santa Clara, CA (US)
Filed by Intel Corporation, Santa Clara, CA (US)
Filed on Aug. 7, 2024, as Appl. No. 18/796,619.
Application 18/796,619 is a continuation of application No. 18/066,436, filed on Dec. 15, 2022, granted, now 12,086,120.
Application 18/066,436 is a continuation of application No. 16/371,342, filed on Apr. 1, 2019, granted, now 11,556,511, issued on Jan. 17, 2023.
Prior Publication US 2025/0036608 A1, Jan. 30, 2025
Int. Cl. G06F 16/22 (2019.01); G06N 20/00 (2019.01); G06T 1/20 (2006.01)
CPC G06F 16/2237 (2019.01) [G06N 20/00 (2019.01); G06T 1/20 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus comprising:
one or more processors including a graphics processor to process data; and
a memory for storage of data, including compressed data;
wherein the one or more processors are to provide for processing of the compressed data, including:
receiving an encoded output vector, the encoded output vector representing a compressed version of an original data structure, the original data structure including a plurality of values;
parsing the encoded output vector to identify a mode, a significance map, and uncompressed data in the encoded output vector, the mode being a most repeated value in the original data structure and the significance map being a structure to indicate locations at which the mode was present in the original data structure; and
decompressing the encoded output vector to generate a data structure, the decompression of the encoded output vector being based at least in part on the identified mode and significance map encoded in the encoded output vector.