# US 12,193,196 B2 · Predictive control using one or more neural networks > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,193,196 B2 - Document ID: US12193196B2 - Status: Granted - Kind: B2 - Application number: 17/023,695 - Filed: 2020-09-17 - Published: 2022-03-17 - Granted: 2025-01-07 - Earliest claimed: 2020-09-17 - Assignees: NVIDIA Corporation - Applicants: Nvidia Corporation - Inventors: Ali Heydari; Vladimir Troy - Web page: https://patentgrep.com/patent/US12193196B2 ## CPC - H05K 7/20836: {Thermal management, e.g. server temperature control} - G05B 13/027: {using neural networks only} - G06N 3/04: Architecture, e.g. interconnection topology - G06N 3/045: Combinations of networks - G06N 3/063: using electronic means - G06N 3/08: Learning methods - G06N 3/088: Non-supervised learning, e.g. competitive learning - G06N 3/09: Supervised learning - H05K 7/20727: {within server blades for removing heat from heat source} - H05K 7/20772: {within server blades for removing heat from heat source} - H05K 7/20736: {within cabinets for removing heat from server blades} - H05K: PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS - G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Systems and methods for cooling a computer environment are disclosed. In at least one embodiment, one or more neural networks can be used to determine one or more temperature control settings associated with one or more servers. ## Claims ### Claim 1 A processor, comprising: one or more circuits to use one or more neural networks associated with one or more different hardware components of a data center to solve a fluid mechanics problem based on an input comprising temperature data and the fluid flow data to predict future fluid flows and future temperatures for a plurality of locations of the data center, the one or more circuits to adjust one or more temperature control settings associated with one or more servers of the data center based on at least one of the future fluid flows or the future temperatures, wherein a first neural network of the one or more networks that is associated with a first type of hardware component is different from a second neural network of the one or more neural networks that is associated with a second type of hardware component. ### Claim 2 (depends on 1) The processor of claim 1, wherein the one or more neural networks are stored on one or more computing boards in one or more temperature control devices associated with the one or more temperature control settings. ### Claim 3 (depends on 1) The processor of claim 1, wherein the one or more temperature control settings are determined for at least one of: individual servers of the one or more servers, individual racks including the one or more servers, or the data center including the one or more servers, and wherein the one or more neural networks can further be used to determine one or more environmental control settings relating to at least one of humidity, fluid flow, or power. ### Claim 4 (depends on 1) The processor of claim 1, wherein the one or more circuits are further to utilize the one or more neural networks over time to determine whether to update the one or more temperature control settings based upon updated predictions generated by the one or more neural networks. ### Claim 5 (depends on 1) The processor of claim 1, wherein the one or more neural networks accept environmental data from at least one of a temperature sensor, a pressure sensor, a flow sensor, a power sensor, a humidity sensor, or a load determination component associated with the one or more servers. ### Claim 6 A system comprising: one or more processors to use one or more neural networks to solve a fluid mechanics problem based on an input comprising temperature data, pressure data and fluid flow data to predict future fluid flows, future pressures and future temperatures for a plurality of locations of a data center, the one or more processors to adjust one or more temperature control settings associated with one or more servers of the data center based on at least one of the future fluid flows, the future pressures or the future temperatures; and a plurality of different data center hardware components, wherein a first neural network of the one or more neural networks that is associated with a first type of data center hardware component is different from a second neural network of the one or more neural networks that is associated with a second type of data center hardware component. ### Claim 7 (depends on 6) The system of claim 6, wherein the one or more neural networks are stored on one or more computing boards in one or more temperature control devices associated with the one or more temperature control settings. ### Claim 8 (depends on 6) The system of claim 6, wherein the one or more temperature control settings are determined for at least one of: individual servers of the one or more servers, individual racks including the one or more servers, or the data center including the one or more servers, and wherein the one or more neural networks can further be used to determine one or more environmental control settings relating to at least one of humidity, fluid flow, or power. ### Claim 9 (depends on 6) The system of claim 6, wherein the one or more processors are further to utilize the one or more neural networks over time to determine whether to update the one or more temperature control settings based upon updated predictions generated by the one or more neural networks. ### Claim 10 (depends on 6) The system of claim 6, wherein the one or more neural networks accept environmental data from at least one of a temperature sensor, a pressure sensor, a flow sensor, a power sensor, a humidity sensor, or a load determination component associated with the one or more servers. ### Claim 11 A method comprising: using one or more neural networks associated with one or more different hardware component of a data center to solve a fluid mechanics problem based on an input comprising temperature data and fluid flow data to predict future fluid flows and future temperatures for a plurality of locations of the data center, wherein a first neural network of the one or more neural networks that is associated with a first type of data center hardware component is different from a second neural network of the one or more neural networks that is associated with a second type of data center hardware component; and adjusting one or more temperature control settings associated with one or more servers of the data center based on at least one of the future fluid flows or the future temperatures. ### Claim 12 (depends on 11) The method of claim 11, wherein the one or more neural networks are stored on one or more computing boards in one or more temperature control devices associated with the one or more temperature control settings. ### Claim 13 (depends on 11) The method of claim 11, wherein the one or more temperature control settings are determined for at least one of: individual servers of the one or more servers, individual racks including the one or more servers, or the data center including the one or more servers, and wherein the one or more neural networks can further be used to determine one or more environmental control settings relating to at least one of humidity, fluid flow, or power. ### Claim 14 (depends on 11) The method of claim 11, further comprising: utilizing the one or more neural networks over time to determine whether to update the one or more temperature control settings based upon updated predictions generated by the one or more neural networks. ### Claim 15 (depends on 11) The method of claim 11, wherein the one or more neural networks accept environmental data from at least one of a temperature sensor, a pressure sensor, a flow sensor, a power sensor, a humidity sensor, or a load determination component associated with the one or more servers. ### Claim 16 A non-transitory machine-readable medium having stored thereon a set of instructions, which if performed by one or more processors, cause the one or more processors to at least: use one or more neural networks associated with one or more different hardware components of a data center to solve a fluid mechanics problem based on an input comprising temperature data and fluid flow data to predict future fluid flows and future temperatures for a plurality of locations of the data center, wherein a first neural network of the one or more neural networks that is associated with a first type of data center hardware component is different from a second neural network of the one or more neural networks that is associated with a second type of data center hardware component; and adjust one or more temperature control settings associated with one or more servers of the data center based on at least one of the future fluid flows or the future temperatures. ### Claim 17 (depends on 16) The non-transitory machine-readable medium of claim 16, wherein the one or more neural networks are stored on one or more computing boards in one or more temperature control devices associated with the one or more temperature control settings. ### Claim 18 (depends on 16) The non-transitory machine-readable medium of claim 16, wherein the one or more temperature control settings are determined for at least one of: individual servers of the one or more servers, individual racks including the one or more servers, or the data center including the one or more servers, and wherein the one or more neural networks can further be used to determine one or more environmental control settings relating to at least one of humidity, fluid flow, or power. ### Claim 19 (depends on 16) The non-transitory machine-readable medium of claim 16, wherein the instructions if performed further cause the one or more processors to: utilize the one or more neural networks over time to determine whether to update the one or more temperature control settings based upon updated predictions generated by the one or more neural networks. ### Claim 20 (depends on 16) The non-transitory machine-readable medium of claim 16, wherein the one or more neural networks accept environmental data from at least one of a temperature sensor, a pressure sensor, a flow sensor, a power sensor, a humidity sensor, or a load determination component associated with the one or more servers. ### Claim 21 A data center cooling system, comprising: one or more cooling systems associated with one or more servers; one or more processors to use one or more neural networks to solve a fluid mechanics problem based on an input comprising the temperature data and the fluid flow data to predict future fluid flows and future temperatures for the plurality of locations of the data center, the one or more processors to adjust one or more temperature control settings for the one or more cooling systems based on at least one of the future fluid flows or the future temperatures; a plurality of different data center hardware components, wherein a first neural network of the one or more neural networks that is associated with a first type of data center hardware component is different from a second neural network of the one or more neural networks that is associated with a second type of data center hardware component; and memory for storing network parameters for the one or more neural networks. ### Claim 22 (depends on 21) The data center cooling system of claim 21, wherein the one or more neural networks are stored on one or more computing boards in one or more temperature control devices associated with the one or more temperature control settings. ### Claim 23 (depends on 21) The data center cooling system of claim 21, wherein the one or more temperature control settings are determined for at least one of: individual servers of the one or more servers, individual racks including the one or more servers, or the data center including the one or more servers, and wherein the one or more neural snetworks can further be used to determine one or more environmental control settings relating to at least one of humidity, fluid flow, or power. ### Claim 24 (depends on 21) The data center cooling system of claim 21, wherein the one or more processors are further to utilize the one or more neural networks over time to determine whether to update the one or more temperature control settings based upon updated predictions generated by the one or more neural networks. ### Claim 25 (depends on 21) The data center cooling system of claim 21, wherein the one or more neural networks accept environmental data from at least one of a temperature sensor, a pressure sensor, a flow sensor, a power sensor, a humidity sensor, or a load determination component associated with the one or more servers.