US 12,131,045 B2
Memory system and operating method thereof
Myung Hyun Rhee, Gyeonggi-do (KR)
Assigned to SK hynix Inc., Gyeonggi-do (KR)
Filed by SK hynix Inc., Gyeonggi-do (KR)
Filed on Jun. 25, 2021, as Appl. No. 17/358,975.
Claims priority of application No. 10-2020-0178442 (KR), filed on Dec. 18, 2020.
Prior Publication US 2022/0197530 A1, Jun. 23, 2022
Int. Cl. G06F 3/06 (2006.01); G06F 18/21 (2023.01); G06F 18/214 (2023.01); G06N 3/04 (2023.01); G06N 3/08 (2023.01)
CPC G06F 3/0644 (2013.01) [G06F 3/0604 (2013.01); G06F 3/0659 (2013.01); G06F 3/0679 (2013.01); G06F 18/214 (2023.01); G06F 18/2163 (2023.01); G06N 3/04 (2013.01); G06N 3/08 (2013.01)] 23 Claims
OG exemplary drawing
 
1. A memory system, comprising:
a plurality of memory devices configured to store pieces of partial data obtained by partitioning an embedding table including pieces of vector information about items of a learning model that have already been acquired; and
a memory controller configured to obtain, in response to a query received from a host, pieces of data corresponding to the query among the pieces of partial data from each of the plurality of memory devices, perform a pooling operation for generating embedding data using the pieces of data that have been obtained, and provide the embedding data to the host,
wherein the query includes a request for the pooling operation, an address of host memory to receive the embedding data, and a physical address corresponding to any one among the plurality of memory devices,
wherein the pieces of vector information indicate embedding vectors corresponding to categorical data,
wherein each of the pieces of partial data is data obtained by partitioning the embedding table by units of a preset number of dimensions of the embedding vectors, and
wherein each of the dimensions corresponds to each of figures included in each of the embedding vectors.