Abstract: Methods, systems and computer program products are provided for improving performance (e.g., reducing power consumption) of a hardware accelerator (e.g., neural processor) comprising hybrid or analog multiply and accumulate (MAC) processing elements (PEs). Selective variation of the precision of an array of MAC PEs may reduce power consumption of a neural processor. Power may be conserved by dynamically controlling the precision of analog to digital (ADC) output bits for one or more MAC PEs. Dynamic control of ADC output bit precision may be based on precision information determined during training and/or post-training (e.g., quantization) of an artificial intelligence (AI) neural network (NN) model implemented by the neural processor. Precision information may include a range of dynamic precision for each of a plurality of nodes of a computation graph for the AI NN model.

Abstract: Methods, systems and computer program products are provided for improving performance (e.g., reducing power consumption) of a hardware accelerator (e.g., neural processor) comprising hybrid or analog multiply and accumulate (MAC) processing elements (PEs). Selective variation of the precision of an array of MAC PEs may reduce power consumption of a neural processor. Power may be conserved by dynamically controlling the precision of analog to digital (ADC) output bits for one or more MAC PEs. Dynamic control of ADC output bit precision may be based on precision information determined during training and/or post-training (e.g., quantization) of an artificial intelligence (AI) neural network (NN) model implemented by the neural processor. Precision information may include a range of dynamic precision for each of a plurality of nodes of a computation graph for the AI NN model.

Abstract: Methods, systems and computer program products are provided for improving performance (e.g., reducing power consumption) of a hardware accelerator (e.g., neural processor) comprising hybrid or analog multiply and accumulate (MAC) processing elements (PEs). Selective variation of the precision of an array of MAC PEs may reduce power consumption of a neural processor. Power may be conserved by dynamically controlling the precision of analog to digital (ADC) output bits for one or more MAC PEs. Dynamic control of ADC output bit precision may be based on precision information determined during training and/or post-training (e.g., quantization) of an artificial intelligence (AI) neural network (NN) model implemented by the neural processor. Precision information may include a range of dynamic precision for each of a plurality of nodes of a computation graph for the AI NN model.

Abstract: Apparatus for determining user operation of an actuator, based on determining a time constant of a state circuit coupled to the actuator.

Abstract: Apparatus for determining user operation of an actuator, based on determining a time constant of a state circuit coupled to the actuator.

Abstract: Apparatus for determining user operation of an actuator, based on determining a time constant of a state circuit coupled to the actuator.

Abstract: Apparatus for determining user operation of an actuator, based on determining a time constant of a state circuit coupled to the actuator.

Abstract: A stylus device for interacting with a computer is disclosed. The stylus device can comprise an operational circuit, an operational switch for activating and deactivating the operational circuit, and a driving and sensing circuit configured for driving a voltage supply to the operational circuit and sensing a state of the operational switch, wherein the driving and the sensing are executed intermittently on an electrical conductor.

Abstract: A stylus device for interacting with a computer is disclosed. The stylus device can comprise an operational circuit, an operational switch for activating and deactivating the operational circuit, and a driving and sensing circuit configured for driving a voltage supply to the operational circuit and sensing a state of the operational switch, wherein the driving and the sensing are executed intermittently on an electrical conductor.

Abstract: A driver circuit for voltage boosting comprises a plurality of circuit cells, each configured to amplify voltage applied thereto; and a plurality of inter-cell switching circuits, arranged to controllably concatenate the cells in series such that, for any pair of adjacent cells, voltage amplified by one cell of the pair is applied, via a respective inter-cell switching circuit, to another cell of the pair. At least one of the inter-cell switching circuits comprises a transistor and a capacitor connected in parallel to each other such that a source-gate voltage of the transistor equals a voltage drop on the capacitor at all times.

Abstract: A driver circuit for voltage boosting comprises a plurality of circuit cells, each configured to amplify voltage applied thereto; and a plurality of inter-cell switching circuits, arranged to controllably concatenate the cells in series such that, for any pair of adjacent cells, voltage amplified by one cell of the pair is applied, via a respective inter-cell switching circuit, to another cell of the pair. At least one of the inter-cell switching circuits comprises a transistor and a capacitor connected in parallel to each other such that a source-gate voltage of the transistor equals a voltage drop on the capacitor at all times.

Abstract: Disclosed are methods, circuits, devices and systems for operating one or more non-volatile memory (NVM) cells within an array of NVM cells. According to embodiments, there may be provided a nonvolatile memory (NVM) device comprising an array of NVM data cells including one or more border/periphery data cells and one or more non-periphery cells. Array control circuitry may be adapted to gauge a state of the one or more periphery data cells differently than non-periphery data cells.

Abstract: A Flash memory array comprises a plurality of Erase Sectors (Esecs) arranged in a plurality of Erase Sector Groups (ESGs), Physical Pages (slices), and Physical Sectors (PSecs), and there is a non-binary number of at least one of the Erase Sector Groups (ESGs), Physical Pages (slices), and Physical Sectors (PSecs). A user address is translated into a physical address using modular arithmetic to determine pointers (ysel, esg, psec) for specifying a given Erase Sector (ESec) within a given Erase Sector Group (ESG); a given Erase Sector Group (ESG) within a given Physical Sector (Psec); and a given Physical Sector (PSec) within the array.

Abstract: Disclosed are methods, circuits, devices and systems for operating one or more non-volatile memory (NVM) cells within an array of NVM cells. According to embodiments, there may be provided a nonvolatile memory (NVM) device comprising an array of NVM data cells including one or more border/periphery data cells and one or more non-periphery cells. Array control circuitry may be adapted to gauge a state of the one or more periphery data cells differently than non-periphery data cells.