Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to adjust vias in integrated circuits (ICs) based on machine learning (ML). An example apparatus computes a dimension by which to extend a via based on at least one of a first metal wire in a first layer of the IC above the via, a via-to-via patterning constraint, or a via-to-metal shorting constraint for a second layer of the IC below the via. The example apparatus also computes a shifted position of the via based on at least one of (a) the dimension or (b) a width and a position of a second metal wire below the via, the width and the position predicted by an ML model. Additionally, the example apparatus adjusts a configuration file corresponding to the IC based on at least one of the dimension or the shifted position of the via.

Abstract: Various embodiments include power management system methods including receiving, at a processor(s), a notification signal triggering the processor(s) to implement power usage mitigation at the processor(s), determining, by the processor(s), a mitigation amount of power rail power by which to mitigate current usage at a power rail based on a use case for the power rail, and implementing power usage mitigation at the processor(s) by the processor(s) sufficient to mitigate power usage at the power rail by the mitigation amount of power rail power. Power usage mitigation may include reducing processor(s) current usage: by a predefined amount; proportional to the amount a power rail current exceeds a power rail current threshold; by the amount of current exceeding a processor current threshold; or by a smallest amount between the amount a power rail current exceeds a power rail current threshold and the processor(s) current exceeds a processor current threshold.

Abstract: A process recipe associated with a substrate at a manufacturing system is identified. A first set of measurements for the substrate is obtained from a substrate measurement subsystem. A second set of measurements for the substrate is obtained from one or more sensors of a chamber of the manufacturing system. A determination is made based on the obtained first set of measurements and the obtained second set of measurements of whether to modify the process recipe by at least one of modifying an operation of the process recipe or generating an instruction to prevent completion of execution of one or more operations of the process recipe.

Abstract: Various embodiments may include methods and systems for power management of multiple chiplets within a system-on-a-chip (SoC). Various systems may include a power management integrated circuit (PMIC) configured to supply power to a first chiplet and a second chiplet across a shared power rail. The first chiplet may be configured to obtain first sensory information throughout the first chiplet. The second chiplet may be configured to obtain second sensory information throughout the second chiplet, and may be configured to transmit a voltage change message to the first chiplet based on the second sensory information. The first chiplet may be configured to transmit a power rail adjustment message to the PMIC based on the first sensory information and the voltage change message. The PMIC may be configured to adjust the voltage of at least one of the first chiplet and the second chiplet.

Abstract: A method for determining whether to modify a manufacturing process recipe is provided. A substrate to be processed at a manufacturing system according to the first process recipe is identified. An instruction to transfer the substrate to a substrate measurement subsystem to obtain a first set of measurements for the substrate is generated. The first set of measurements for the substrate is received from the substrate measurement subsystem. An instruction to transfer the substrate from the substrate measurement subsystem to a processing chamber is generated. A second set of measurements for the substrate is received from one or more sensors of the processing chamber. A first mapping between the first set of measurements and the second set of measurements for the substrate is generated. The first set of measurements mapped to the second set of measurements for the substrate is stored.

Abstract: Task scheduling in a computing device may be based in part on voltage regulator efficiency. For an additional task to be scheduled, multiple task scheduling cases may be determined that represent execution of the additional task on each of a number of processors concurrently with one or more other tasks executing among the processors. For each task scheduling case, a regulator input power level for a voltage regulator may be determined based on a performance level indication associated with the additional task, the one or more other tasks executing on the processors, and the efficiency level of each voltage regulator. For each task scheduling case, a total regulator input power level may be determined by summing the regulator input power levels for all voltage regulators. The additional task may be executed on a processor associated with a task scheduling case for which total regulator input power is lowest.

Abstract: A method for mitigating loads acting on a rotor blade of a wind turbine includes receiving a plurality of loading signals and determining at least one load acting on the rotor blade based on the loading signals. Further, the method includes determining a type of the load(s) acting on the rotor blade. Moreover, the method includes comparing the load(s) to a loading threshold, such as an extreme loading threshold. In addition, the method includes implementing a control scheme when the load(s) exceeds the loading threshold. More specifically, the control scheme includes providing a first pitching mode for reducing a first type of load, providing a different, second pitching mode for reducing a different, second type of load, and coordinating the first and second pitching modes based on the type of the at least one load to mitigate the loads acting on the rotor blade.

Abstract: A system includes a processing device, operatively coupled to the memory device, to perform operations comprising obtaining a plurality of sensor values associated with a deposition process performed, according to a recipe, in a process chamber to deposit film on a surface of a substrate; generating a manufacturing data graph based on the plurality of sensor values; receiving, via a user interface, a selection of a data point on the manufacturing graph; receiving failure data associated with the data point; and storing, in a data structure, the failure data to be accessible via the user interface presenting the manufacturing data graph.

Abstract: Spectral data associated with a first prior substrate and/or a second prior substrate is obtained. A metrology measurement value associated with the first portion of the first prior substrate is determined based on one or more metrology measurement values measured for at least one of a second portion of the first prior substrate or a third portion of a second prior substrate. Training data for training a machine learning model to predict metrology measurement values of a current substrate is generated. Generating the training data includes generating a first training input including the spectral data associated with the first prior substrate and generating a first target output for the first training input, the first target output including the determined metrology measurement value associated with the first portion of the first prior substrate. The training data is provided to train the machine learning model.

Abstract: Various embodiments may include methods and systems for power management of multiple chiplets within a system-on-a-chip (SoC). Various systems may include a power management integrated circuit (PMIC) configured to supply power to a first chiplet and a second chiplet across a shared power rail. The first chiplet may be configured to obtain first sensory information throughout the first chiplet. The second chiplet may be configured to obtain second sensory information throughout the second chiplet, and may be configured to transmit a voltage change message to the first chiplet based on the second sensory information. The first chiplet may be configured to transmit a power rail adjustment message to the PMIC based on the first sensory information and the voltage change message. The PMIC may be configured to adjust the voltage of at least one of the first chiplet and the second chiplet.

Abstract: A method for reducing extreme loads acting on a component of a wind turbine includes measuring, via one or more sensors, a plurality of operating parameters of the wind turbine. Further, the method includes predicting at least one blade moment of at least one rotor blade of the wind turbine based on the plurality of operating parameters. The method also includes predicting a load and an associated load angle of the at least one rotor blade as a function of the at least one blade moment. Moreover, the method includes predicting a pitch angle of the at least one rotor blade of the wind turbine. In addition, the method includes generating a load envelope for the component that comprises at least one load value for the pitch angle and the load angle. Thus, the method includes implementing a control action when the load is outside of the load envelope.

Abstract: Task scheduling in a computing device may be based in part on voltage regulator efficiency. For an additional task to be scheduled, multiple task scheduling cases may be determined that represent execution of the additional task on each of a number of processors concurrently with one or more other tasks executing among the processors. For each task scheduling case, a regulator input power level for a voltage regulator may be determined based on a performance level indication associated with the additional task, the one or more other tasks executing on the processors, and the efficiency level of each voltage regulator. For each task scheduling case, a total regulator input power level may be determined by summing the regulator input power levels for all voltage regulators. The additional task may be executed on a processor associated with a task scheduling case for which total regulator input power is lowest.

Abstract: Systems, methods, and apparatus for power management are disclosed. A power management integrated circuit has a bus interface circuit configured to couple the power management integrated circuit to a shared communication bus, one or more regulator circuits configured to provide current to a managed device, and a controller. The controller is configured to determine that current consumption by the managed device exceeds a threshold level, generate an extended current level message to be transmitted over the shared communication bus to the managed device and transmit a time value with the extended current level message, the time value indicative of an elapsed time between generation of the extended current level message and start of transmission of the extended current level message.

Abstract: Methods and systems for detecting and correcting substrate process drift using machine learning are provided. Data associated with processing each of a first set of substrates at a manufacturing system according to a process recipe is provided as input to a trained machine learning model. One or more outputs are obtained from the trained machine learning model. An amount of drift of a first set of metrology measurement values for the first set of substrates from a target metrology measurement value is determined from the one or more outputs. Process recipe modification identifying one or more modifications to the process recipe is also determined. For each modification, an indication of a level of confidence that a respective modification to the process recipe satisfies a drift criterion for a second set of substrates is determined.

Abstract: A method for training a machine learning model to predict metrology measurements of a current substrate being processed at a manufacturing system is provided. Training data for the machine learning model is generated. A first training input including historical spectral data and/or historical non-spectral data associated with a surface of a prior substrate previously processed at the manufacturing system is generated. A first target output for the first training input is generated. The first target output includes historical metrology measurements associated with the prior substrate previously processed at the manufacturing system. Data is provided to train the machine learning model on (i) a set of training inputs including the first training input, and (ii) a set of target outputs including a first target output.

Abstract: A method for a substrate measurement subsystem is provided. An indication is received that a substrate being processed at a manufacturing system has been loaded into a substrate measurement subsystem. First positional data of the substrate within the substrate measurement subsystem is determined. One or more portions of the substrate to be measured by one or more sensing components of the substrate measurement subsystem are determined based on the first positional data of the substrate and a process recipe for the substrate. Measurements of each of the determined portions of the substrate are obtained by one or more sensing components of the substrate measurement subsystem. The obtained measurements of each of the determined portions of the substrate are transmitted to a system controller.

Abstract: A method for determining whether to modify a manufacturing process recipe is provided. A substrate to be processed at a manufacturing system according to the first process recipe is identified. An instruction to transfer the substrate to a substrate measurement subsystem to obtain a first set of measurements for the substrate is generated. The first set of measurements for the substrate is received from the substrate measurement subsystem. An instruction to transfer the substrate from the substrate measurement subsystem to a processing chamber is generated. A second set of measurements for the substrate is received from one or more sensors of the processing chamber. A first mapping between the first set of measurements and the second set of measurements for the substrate is generated. The first set of measurements mapped to the second set of measurements for the substrate is stored.

Abstract: A method for mitigating loads acting on a rotor blade of a wind turbine includes receiving a plurality of loading signals and determining at least one load acting on the rotor blade based on the loading signals. Further, the method includes determining a type of the load(s) acting on the rotor blade. Moreover, the method includes comparing the load(s) to a loading threshold, such as an extreme loading threshold. In addition, the method includes implementing a control scheme when the load(s) exceeds the loading threshold. More specifically, the control scheme includes providing a first pitching mode for reducing a first type of load, providing a different, second pitching mode for reducing a different, second type of load, and coordinating the first and second pitching modes based on the type of the at least one load to mitigate the loads acting on the rotor blade.

Abstract: A method for reducing extreme loads acting on a component of a wind turbine includes measuring, via one or more sensors, a plurality of operating parameters of the wind turbine. Further, the method includes predicting at least one blade moment of at least one rotor blade of the wind turbine based on the plurality of operating parameters. The method also includes predicting a load and an associated load angle of the at least one rotor blade as a function of the at least one blade moment. Moreover, the method includes predicting a pitch angle of the at least one rotor blade of the wind turbine. In addition, the method includes generating a load envelope for the component that comprises at least one load value for the pitch angle and the load angle. Thus, the method includes implementing a control action when the load is outside of the load envelope.