Abstract: The present disclosure relates to a method for training a neural network model performed at an electronic device. The method includes: performing initial training by using a first training sample set to obtain an initial neural network model; performing a prediction on a second training sample set by using the initial neural network model to obtain a prediction result of each of training samples in the second training sample set; determining a plurality of preferred samples from the second training sample set based on the prediction results; adding the plurality of preferred samples that are annotated to the first training sample set to obtain an expanded first training sample set; updating training of the initial neural network model by using the expanded first training sample set to obtain an updated neural network model until a training ending condition is satisfied.

Abstract: A host system for executing an application on first and/or second reconfigurable processors is presented. The host system is operatively coupled to the first and second reconfigurable processors, whereby the first reconfigurable processors have a first architecture, and the second reconfigurable processors have a second architecture that is different than the first architecture. The host system allocates reconfigurable processors of the first and/or second reconfigurable processors for executing the application and includes an auto-discovery module that is configured to determine whether the allocated reconfigurable processors include at least one of the first reconfigurable processors.

Abstract: A device includes a first memory bank and a second memory bank. The first memory bank is configured to operate according to a write data signal and a first global write signal associated with a first clock signal. The second memory bank is configured to operate according to the write data signal and a second global write signal associated with a second clock signal. One of the first clock signal and the second clock signal is in oscillation when another one of the first clock signal and the second clock signal is in suspension.

Abstract: A single-ended to a differential signal converter (converter) includes, in part, first, second, and third inverting elements, each having a first size, and coupled in series to form a chain of inverting elements. The converter further includes a fourth inverting element of a second size and coupled to the input of the first inverting element, a fifth inverting element of a third size and coupled to an output terminal of the first inverting element, a sixth inverting element of the third size and coupled to an output of the second inverting element, and a seventh inverting element of the second size and coupled to the output of the third inverting element. The outputs of the fourth and sixth inverting elements form a first one of the differential signals. The outputs of the fifth and seventh inverting elements form a second one of the differential signals.

Abstract: A method includes: receiving a data signal (DQ) and a clock signal (DQS); generating an indicator signal by: delaying the data signal (DQ) to generate a delayed data signal (DQ?); sampling the data signal (DQ) and the delayed data signal (DQ?) using an edge of the clock signal (DQS) to generate a first sampled value and a second sampled value; and generating the indicator signal based on the first sampled value and the second sampled value; and adjusting one or more of a DQ adjustable delay line associated with the data signal (DQ) and a DQS adjustable delay line associated with the clock signal (DQS) based on the indicator signal to synchronize the data signal (DQ) and the clock signal (DQS).

Abstract: A system for executing an application on a pool of reconfigurable processors with first and second reconfigurable processors having first and second architectures that are different from each other is presented. The system comprises an archive of configuration files with first and second configuration files for executing the application on the first and second reconfigurable processors, respectively, and a host system that is operatively coupled to the first and second reconfigurable processors. The host system comprises a runtime processor that allocates reconfigurable processors for executing the application and an auto-discovery module that is configured to perform discovery of whether the reconfigurable processors include at least one of the first reconfigurable processors and whether the reconfigurable processors include at least one of the second reconfigurable processors.

Abstract: In a method an orchestrator of a computing system determines that results of Machine Learning model computations are available and dispatches a worker to perform model computations that include computing gradients of the results. The orchestrator determines that a set of gradients of the results is available and dispatches a gradient worker to compute a sum of the gradients. The orchestrator determines that a second set of gradients of the results is available and dispatches a second gradient worker to compute a sum of the second set of gradients. The orchestrator determines that the sums of the first and second gradients are available and dispatches a third gradient worker to compute synchronized gradients. The gradient workers compute the sums and synchronized gradients concurrent with training workers computing additional model computations results and/or gradients. A computer program product can include the method and a computing system can include the orchestrator.

Abstract: An apparatus is provided, where the apparatus includes a plurality of components; a first circuitry to generate a clock signal, and to supply the clock signal to the plurality of components; a second circuitry to estimate, for each of two or more components of the plurality of components, a corresponding duty cycle of the clock signal received at the corresponding component, wherein two or more duty cycles corresponding to the two or more components are determined; a third circuitry to determine an average of the two or more duty cycles; and a fourth circuitry to correct a duty cycle of the clock signal generated by the first circuitry, based at least in part on the average.

Abstract: A system for a data-parallel execution of at least two implementations of an application on reconfigurable processors with different layouts is presented. The system comprises a pool of reconfigurable data flow resources with data transfer resources that interconnect first and second reconfigurable processors having first and second layouts that impose respective first and second constraints for the data-parallel execution of the application. The system further comprises an archive of configuration files and a host system that is operatively coupled to the first and second reconfigurable processors. The host system comprises first and second compilers that generate for the application, based on the respective first and second constraints, first and second configuration files that are stored in the archive of configuration files and adapted to be executed data-parallel compatible on respective first and second reconfigurable processors.

Abstract: A system for executing an application on a pool of reconfigurable processors with first and second reconfigurable processors having first and second architectures that are different from each other is presented. The system comprises an archive of configuration files with first and second configuration files for executing the application on the first and second reconfigurable processors, respectively, and a host system that is operatively coupled to the first and second reconfigurable processors. The host system comprises a runtime processor that allocates reconfigurable processors for executing the application and an auto-discovery module that is configured to perform discovery of whether the reconfigurable processors include at least one of the first reconfigurable processors and whether the reconfigurable processors include at least one of the second reconfigurable processors.

Abstract: A head-mounted display device includes a see-though display providing both eyes of a user with a view of a physical object, a processor, and a non-volatile storage device holding instructions executable by the processor to: display an image that corresponds to the physical object to a first eye of the user at an offset to the physical object; display blocking light to a second eye of the user; in response to alignment user input, move a position of the image relative to the physical object; in response to completion user input, determine the inter-pupillary distance of the user; and calibrate the head-mounted display device based on the inter-pupillary distance.

Abstract: Duty cycle adjustment circuitry includes a first stage, a second stage, and decoder circuitry. The first stage includes a first strength tuning circuit having first inverter branches, and a first fine tuning circuit having second inverter branches. The first strength tuning circuit and the first fine tuning circuit are coupled in parallel. The second stage includes a second strength tuning circuit having third inverter branches, and a second fine tuning circuit having fourth inverter branches. The second strength tuning circuit and the second fine tuning circuit are coupled in parallel. Further, the second stage is electrically coupled to the first stage. The decoder circuitry is electrically coupled to the first stage and the second stage. The decoder circuitry controls the first strength tuning circuit independently from the first fine tuning circuit to adjust the duty cycle of an input signal received by the duty cycle adjustment circuitry.

Abstract: Duty cycle adjustment circuitry includes a first stage, a second stage, and decoder circuitry. The first stage includes a first strength tuning circuit having first inverter branches, and a first fine tuning circuit having second inverter branches. The first strength tuning circuit and the first fine tuning circuit are coupled in parallel. The second stage includes a second strength tuning circuit having third inverter branches, and a second fine tuning circuit having fourth inverter branches. The second strength tuning circuit and the second fine tuning circuit are coupled in parallel. Further, the second stage is electrically coupled to the first stage. The decoder circuitry is electrically coupled to the first stage and the second stage. The decoder circuitry controls the first strength tuning circuit independently from the first fine tuning circuit to adjust the duty cycle of an input signal received by the duty cycle adjustment circuitry.

Abstract: A head-mounted display device includes a see-though display providing both eyes of a user with a view of a physical object, a processor, and a non-volatile storage device holding instructions executable by the processor to: display an image that corresponds to the physical object to a first eye of the user at an offset to the physical object; display blocking light to a second eye of the user; in response to alignment user input, move a position of the image relative to the physical object; in response to completion user input, determine the inter-pupillary distance of the user; and calibrate the head-mounted display device based on the inter-pupillary distance.

Abstract: A head-mounted display device is disclosed that includes an at least partially see-though display, a processor, and a non-volatile storage device holding instructions executable by the processor to: select an image that corresponds to a physical object viewable by the user; display the image at a perceived offset to the physical object; in response to alignment user input, move a perceived position of the image relative to the physical object; output an instruction to provide completion user input when the image appears to align with the physical object; when the completion user input is received, determine the inter-pupillary distance of the user; and calibrate the head mounted display device based on the inter-pupillary distance.

Abstract: An apparatus is provided, where the apparatus includes a plurality of components; a first circuitry to generate a clock signal, and to supply the clock signal to the plurality of components; a second circuitry to estimate, for each of two or more components of the plurality of components, a corresponding duty cycle of the clock signal received at the corresponding component, wherein two or more duty cycles corresponding to the two or more components are determined; a third circuitry to determine an average of the two or more duty cycles; and a fourth circuitry to correct a duty cycle of the clock signal generated by the first circuitry, based at least in part on the average.

Abstract: An apparatus is provided, where the apparatus includes a plurality of components; a first circuitry to generate a clock signal, and to supply the clock signal to the plurality of components; a second circuitry to estimate, for each of two or more components of the plurality of components, a corresponding duty cycle of the clock signal received at the corresponding component, wherein two or more duty cycles corresponding to the two or more components are determined; a third circuitry to determine an average of the two or more duty cycles; and a fourth circuitry to correct a duty cycle of the clock signal generated by the first circuitry, based at least in part on the average.

Abstract: An apparatus is provided, where the apparatus includes a plurality of components; a first circuitry to generate a clock signal, and to supply the clock signal to the plurality of components; a second circuitry to estimate, for each of two or more components of the plurality of components, a corresponding duty cycle of the clock signal received at the corresponding component, wherein two or more duty cycles corresponding to the two or more components are determined; a third circuitry to determine an average of the two or more duty cycles; and a fourth circuitry to correct a duty cycle of the clock signal generated by the first circuitry, based at least in part on the average.

Abstract: An embodiment of a receiver apparatus may include high pass components to pass high frequency components of an input signal, low pass components to pass low frequency components of the input signal, and an amplifier communicatively coupled to the high pass components and the low pass components to amplify respective signals passed by the high pass components and the low pass components, wherein the low pass components include a level shifter to shift a common mode voltage level of the input signal to a switch threshold voltage for the amplifier in accordance with at least two different types of memory devices. Other embodiments are disclosed and claimed.

Abstract: An embodiment of a receiver apparatus may include high pass components to pass high frequency components of an input signal, low pass components to pass low frequency components of the input signal, and an amplifier communicatively coupled to the high pass components and the low pass components to amplify respective signals passed by the high pass components and the low pass components, wherein the low pass components include a level shifter to shift a common mode voltage level of the input signal to a switch threshold voltage for the amplifier in accordance with at least two different types of memory devices. Other embodiments are disclosed and claimed.