Abstract: A memory cooler includes a unitary thermal transfer device and a pair of endcaps. The unitary thermal transfer device includes heat transfer tubes, a first end block, a second end block, an inlet chamber, an outlet chamber, an inlet, and an outlet. The first and second end blocks are structurally integrated with each of the heat transfer tubes. The inlet and outlet chambers are partially defined by either the first end block or the second end block. Each of the inlet and outlet chambers are fluidly coupled with the respective liquid flow channel of the at least one heat transfer tube. The respective flow channels to which the inlet and outlet chambers are coupled may be the same or different to define either a direct or a serpentine flow path. Each endcap is affixed to a respective one of the first end block and the second end block to define, in conjunction with the first end block and second end block, the inlet chamber and the outlet chamber.

Abstract: In exemplary aspects of optimizing data centers, historical data corresponding to a data center is collected. The data center includes a plurality of systems. A data center representation is generated. The data center representation can be one or more of a schematic and a collection of data from among the historical data. The data center representation is encoded into a neural network model. The neural network model is trained using at least a portion of the historical data. The trained model is deployed using a first set of inputs, causing the model to generate one or more output values for managing or optimizing the data center with respect to design and risk aspects.

Abstract: Hose barb fittings and apparatuses described herein provide increased fluid-flow rates for cooling loops used for thermal control in computer system. A hose barb fitting comprises a fluid-flow passage that extends through the hose barb fitting from a first opening to a second opening. The ratio of the cross-sectional area of the fluid-flow passage to the cross-sectional area of the hose barb fitting is between 0.4 and 0.7, inclusive. When the hose barb fitting is fully seated within a housing structure, a specialized gasket acts as both a radial seal and a face seal. Also, a flange extending from the housing structure engages with a flange extending from the hose barb fitting to prevent the hose barb fitting from being unseated.

Abstract: Systems and methods are provided for improving statistical and machine learning drift detection models that monitor computing health of a data center environment. For example, the system can receive streams of sensor data from a plurality of sensors in a data center; clean the streams of sensor data; generate, using a machine learning (ML) model, an anomaly score and a dynamic threshold value based on the cleaned streams of sensor data; determine, using the ML model and based on the anomaly score and the dynamic threshold value, a correctness indicator for a first sensor in the plurality of sensors; and using the correctness indicator, correct the first sensor.

Abstract: A compute device includes a printed circuit board, at least three compute subassemblies disposed on the printed circuit board, and a liquid loop. The compute subassemblies disposed on the printed circuit board and each of the three compute subassemblies includes a thermal control plate defining a respective internal conduit therethrough. A temperature controlled liquid circuit circulates through the liquid loop through to control the temperature of each of compute subassemblies in series during operation, the liquid loop including each of the internal conduits in each of the thermal control plates in each of the compute subassemblies.

Abstract: Multi-metric artificial intelligence (AI)/machine learning (ML) models for detection of anomalous behavior of a machine/system are disclosed. The multi-metric AI/ML models are configured to detect anomalous behavior of systems having multiple sensors that measure correlated sensor metrics such as coolant distribution units (CDUs). The multi-metric AI/ML models perform the anomalous system behavior detection in a manner that enables both a reduction in the amount of sensor instrumentation needed to monitor the system's operational behavior as well as a corresponding reduction in the complexity of the firmware that controls the sensor instrumentation. As such, AI-enabled systems and corresponding methods for anomalous behavior detection disclosed herein offer a technical solution to the technical problem of increased failure rates of existing multi-sensor systems, which is caused by the presence of redundant sensor instrumentation that necessitates complex firmware for controlling the sensor instrumentation.

Abstract: A compute device includes a printed circuit board, at least three compute subassemblies disposed on the printed circuit board, and a liquid loop. The compute subassemblies disposed on the printed circuit board and each of the three compute subassemblies includes a thermal control plate defining a respective internal conduit therethrough. A temperature controlled liquid circuit circulates through the liquid loop through to control the temperature of each of compute subassemblies in series during operation, the liquid loop including each of the internal conduits in each of the thermal control plates in each of the compute subassemblies.

Abstract: A circuit board cooling apparatus is disclosed having a heat spreader device to be thermally coupled with a surface of the circuit board to be cooled. Also, a conforming heat transfer device is disclosed that is thermally and physically coupled with the heat spreader device to conform to a surface contour of the heat spreader device on a first side of the heat transfer device. The cooling apparatus also includes a heat transport device physically attached and thermally coupled with a second side of the heat transfer device.

Abstract: Example implementations relate to a method and system for pumping fluid with a pump assembly to cool waste-heat producing system of an apparatus. The pump assembly includes a first housing, a rotor assembly, a stator assembly, a second housing, an outlet connection port coupled to the second housing, and a guide vane insert. The first housing includes a first bore and first inlet guide vanes (IGVs) coupled to inlet of the first bore. The rotor assembly including an impeller, is disposed in the first bore. The stator assembly including a coil, is mounted around a portion of the first housing. The second housing has a second bore fluidically coupled to the first bore, and first outlet guide vanes (OGVs) coupled at an entrance of the second bore. The guide vane insert including second OGVs, is disposed within the outlet connection port and fluidically coupled to outlet of the second bore.

Abstract: Example implementations relate to a method and system for cooling a compute node tray including a board, a plurality of first and second devices, a cooling assembly, and a support structure assembly. The cooling assembly includes a supply section, a return section, and an intermediate section coupled to the supply and return sections. The supply section includes a first conduit extending along a perimeter of the board and forming a thermal contact with the first devices. The intermediate section includes a plurality of cold plates and a plurality of second conduits forming the thermal contact with the second devices. The second conduits extends parallel to one another. The return section includes a third conduit extending parallel to a portion of the first conduit. The support structure encompasses the board and forms the thermal contact with a portion of the board, a plurality of third devices, and the cooling assembly.

Abstract: Example implementations relate to a method and system for cooling a compute node tray including a board, a plurality of first and second devices, a cooling assembly, and a support structure assembly. The cooling assembly includes a supply section, a return section, and an intermediate section coupled to the supply and return sections. The supply section includes a first conduit extending along a perimeter of the board and forming a thermal contact with the first devices. The intermediate section includes a plurality of cold plates and a plurality of second conduits forming the thermal contact with the second devices. The second conduits extends parallel to one another. The return section includes a third conduit extending parallel to a portion of the first conduit. The support structure encompasses the board and forms the thermal contact with a portion of the board, a plurality of third devices, and the cooling assembly.

Abstract: An apparatus is provided herein. The apparatus includes a sensor module and a control module. The sensor module to receive a measured environmental condition. The control module to use the measured environmental condition to determine a fluid temperature to cool a first set of components and determine an air temperature to cool a second set of components.

Abstract: Systems and devices for cooling servers are provided. In one aspect, a cooling device includes a first axial pump including a body having an impeller, the first axial pump coupled to a first pump housing, the first axial pump housing coupled to a chassis, a rack, a row of racks, or one or more racks of the row of racks that are housing one or more servers. The cooling device also includes an inlet pipe coupled to an inlet of the first pump housing, the inlet pipe supplying cooling fluid to the first axial pump. The cooling device also includes an outlet pipe having an outlet coupled to the first pump housing, the outlet pipe receiving the cooling fluid from the first axial pump.

Abstract: In exemplary aspects of optimizing data centers, historical data corresponding to a data center is collected. The data center includes a plurality of systems. A data center representation is generated. The data center representation can be one or more of a schematic and a collection of data from among the historical data. The data center representation is encoded into a neural network model. The neural network model is trained using at least a portion of the historical data. The trained model is deployed using a first set of inputs, causing the model to generate one or more output values for managing or optimizing the data center using supplemental indicators.

Abstract: An example device includes processing circuitry and a memory. The memory includes instructions that cause the device to perform various functions. The functions include receiving datastreams from a plurality of sensors of a high performance computing system, classifying each datastream of the each sensor to one of a plurality of datastream models, selecting an anomaly detection algorithm from a plurality of anomaly detection algorithms for each datastream, determining parameters of the each anomaly detection algorithm, determining an anomaly threshold for each datastream, and generating an indication that the sensor associated with the datastream is acting anomalously.

Abstract: An axial flow pump comprises a housing having a bore, a cylindrical permanent magnet within the housing bore, and at least one impeller inside the permanent magnet and adapted to cause fluid to flow within the cylindrical permanent magnet. The permanent magnet is configured to rotate in the bore around its longitudinal axis. A motor lamination stack surrounds the cylindrical permanent magnet. The motor lamination stack is formed substantially as a cuboid, extending in a width and height direction perpendicular to the axis of rotation of the cylindrical permanent magnet. The diameter of the bore is at least 80% of the dimension of the cuboid in the height direction. A coil of the motor lamination is configured to be energized and to create a rotating magnetic field in the lamination stack to rotate the cylindrical permanent magnet around its longitudinal axis. The coil is disposed to one or both sides of the pump.

Abstract: In exemplary aspects of managing, monitoring and maintaining computing systems and devices such as edge data centers (EDCs), probabilistic models such as dynamic Bayesian networks (DBNs) are generated. The DBNs can define individual and collective systems such as EDCs. The DBNs are built by generating or estimating the model structure and model parameters. The model can be deployed, for instance, to identify actual or potentially anomalous behavior within the individual or collective systems defined by the model. The model can also be deployed to predict anomalous behavior. Based on the results of the model, corrective measures can be taken to remedy the anomalies, and/or to optimize the impact therefrom.

Abstract: Methods and systems for creating cold plates are disclosed. For example, in one example method for manufacturing a cold plate made of a thermally-conductive plastic includes performing a first injection-molding process using the thermally-conductive plastic to produce a first unitary body, the first unitary body including one or more elongated sections forming a unitary body, performing a second injection-molding process using the thermally-conductive plastic to produce a second unitary body, the second unitary body incorporating the first unitary body so as to cover a majority of the first unitary body and form a respective coolant pipe body corresponding to each elongated section, and performing a machining process on the second unitary body so as to create a conduit in each respective coolant pipe body suitable for a fluid to pass through to create respective coolant pipes.

Abstract: In one implementation, a system for scalable coolant distribution unit includes a parameters engine to determine real time pump parameters of a first CDU, wherein the real time pump parameters correspond to a functionality of the first CDU, a pump engine to alter pump parameters of the first CDU based on the real time pump parameters, a communication engine to send the altered pump parameters of the first CDU to a second CDU, the pump engine to alter pump parameters of the second CDU based on the altered pump parameters of the first CDU and determined real time pump parameters of the second CDU, a functionality engine to determine a functionality of the first CDU and the second CDU based on the real time pump parameters and altered pump parameters.

Abstract: In some examples, a system may include a plurality of cooling distribution units (CDUs) to control cooling of a plurality of computing devices based in part on average differential pressure between a supply and return for each of the plurality of CDUs, temperature of coolant in a loop flowing between each of the plurality of CDUs and computing devices, and facility valve position for each of the plurality of CDUs, wherein one CDU of the plurality of the CDUs is nominated lead CDU and broadcasts flow rate to the plurality of CDUs to follow.