US 11,864,359 B2
Intelligent threshold leak remediaton of datacenter cooling systems
Ali Heydari, Albany, CA (US); Mike Sabotta, Cypress, TX (US); and Harold Miyamura, San Jose, CA (US)
Assigned to Nvidia Corporation, Santa Clara, CA (US)
Filed by Nvidia Corporation, Santa Clara, CA (US)
Filed on Aug. 27, 2020, as Appl. No. 17/005,017.
Prior Publication US 2022/0071049 A1, Mar. 3, 2022
Int. Cl. H05K 7/20 (2006.01)
CPC H05K 7/2079 (2013.01) [H05K 7/20281 (2013.01); H05K 7/20836 (2013.01)] 10 Claims
OG exemplary drawing
 
1. At least one processor for a liquid cooling system, comprising:
at least one logic unit to train one or more neural networks having hidden layers of neurons to evaluate parameter values associated with at least one parameter of the liquid cooling system, the evaluation performed using the parameter values and using prior associated flow rates for a coolant and prior associated power states for at least one computing component, the evaluation to determine a moving range for the at least one parameter that is outside a normal threshold and that is indicative of a threshold leak of the coolant has occurred, the threshold leak associated with the at least one computing component functioning within a normal temperature threshold and receiving the coolant, and the one or more neural network to provide outputs associated with a first change in a power state for the at least one computing component to reduce reliance on the coolant and associated with a second change to a flow of the coolant to the at least one computing component.