Predicate |
Object |
assignee |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentassignee/MD5_51d028c578ae85cb937b5b34a5129fbc |
classificationCPCAdditional |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-088 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N5-046 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-084 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-20084 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-047 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-20081 |
classificationCPCInventive |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F9-3802 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-04 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-044 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-045 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-048 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T1-20 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-063 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-08 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-084 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F9-3804 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F9-3887 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T5-002 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F9-5027 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N20-10 |
classificationIPCInventive |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06F9-50 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06T1-40 |
filingDate |
2020-11-09-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
inventor |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_e4cece0e6a6669a7b56ef37e28625f2a http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_e1036805b0a747515e5e68aae54e92a6 http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_c13ba7ac9ecadfb503b293e9b7e012e4 http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_42dfe0712899fc997a93b81d1f021ef4 http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_b274fa9128a16d311b5198fa9e317870 |
publicationDate |
2021-05-12-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber |
DE-102020129409-A1 |
titleOfInvention |
DYNAMIC DIVISION OF ACTIVATIONS AND KERNELS TO IMPROVE STORAGE EFFICIENCY |
abstract |
Embodiments generally focus on dynamically dividing activations and kernels to improve memory efficiency. One embodiment of a method in a computation engine, performing machine learning, comprises the following: receiving, by a folding layer of a folding neural network (CNN) implemented on the computation engine, from a plurality of activation groups contained in the input data, the folding layer one or more kernel groups and the one or more kernel groups each comprise a plurality of kernels; Determining a plurality of memory efficiency metrics based on the number of activation groups of the plurality of activation groups and the number of kernels of the plurality of kernels; Selecting a first optimal number of activation groups and a second optimal number of kernels associated with an optimal memory efficiency metric in the plurality of memory efficiency metrics; and performing a convolving operation on the input data based on the first optimal number and the second optimal number. |
priorityDate |
2019-11-07-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type |
http://data.epo.org/linked-data/def/patent/Publication |