Abstract: A processor set has an associated cache memory. A prefetch unit of the processor set generates a prefetch request. The prefetch unit determines whether the prefetch request is likely to miss in the cache memory. The prefetch unit issues the prefetch request based on determining that the prefetch request is likely to miss in the cache memory and otherwise discards the prefetch request. The prefetch request can be a data prefetch request or an instruction prefetch request.

Abstract: A method for performing integrated circuit testing is presented. A compactor is generated using explicit expander graph construction. A decompressor and the generated compactor are inserted within a device-under-tested (DUT). After feeding a compressed test pattern into the DUT, the compressed test pattern is expanded via the decompressor into internal scan-ins within the DUT. Subsequently, the internal scan-out data is compressed into compressed output vectors using the generated compactor. After shifting out from the DUT, the compressed output vectors are compared to a set of desired outputs in order to determine a pass/fail status of the DUT.

Abstract: According to one embodiment, a method, computer system, and computer program product for mixed reality is provided. The present invention may include receiving 3D hand keypoints (keypoints) of a user's visible hand joints from the user's capturable hand, and visible hand joints, if any, from the user's uncapturable hand; using random noise sampled with a unit normal distribution as initial keypoints for the uncapturable hand joints from the user's uncapturable hand; inputting the received and the initial keypoints, in a preset order, into a trained 3D hand joint generative model; performing an iterative refinement of the uncapturable hand joints from the user's uncapturable hand using the trained 3D hand joint generative model; identifying whether generated keypoints of the user's uncapturable hand are synchronized with the keypoints of the user's capturable hand; and rendering the generated 3D keypoints for the user's uncapturable hand joints using a 3D virtual hand modeler.

Abstract: Provided are a tape head controller, tape drive, and method for simultaneously writing data to two sets of channels to increase channel transfer rate. A tape head module having a first and second sets of read/write elements is positioned to have the first set of read/write elements on a selected wrap number in a first n channels in a first data band and the second set of read/write elements on the selected wrap number in a second n channels in a second data band. The first set and the second set of read/write elements are controlled to simultaneously perform read/write operations with respect to the selected wrap number in the first n channels and the second n channels, respectively. A combination of the first n channels and the second n channels is presented as a 2n-channel layout with a plurality of wraps in each of n channels.

Abstract: A method, by an intelligent access control system, of implementing access control to a controlled area having a plurality access areas divided by at least one access point includes the following operations. A security token is generated. The security token is transmitted to a client module executing on a client device associated with a first user, and the client module is caused to pair with an access control device associated with the first user. The client module is caused to transfer the security token to the access control device. A physical location of the client device within the controlled area is monitored. A determination is made, by an access control system and based upon the physical location, that the first user is to be revalidated. The client module causes, based upon the determining, the client device to generate an alert.

Abstract: A computer-implemented method to initiate unplanned interactions for remote workers. The method includes identifying a plurality of user accounts, where each user account is associated with a communication device and a physical location. The method further includes determining a location of each communication device within each physical location. The method also includes generating, for each user account, a collaboration score, where the collaboration scores are based on the location of each communication device. The method includes determining a first collaboration score of a first user account and a second collaboration score of a second user account are above a collaboration threshold. The method also includes initiating, based on the first collaboration score and the second collaboration score being above the threshold, a communication channel between a first communication device and a second communication device.

Abstract: In an approach, a processor receives a plurality of first operand values, where the first operand values are integer values. A processor adds, using binary addition, the plurality of first operand values resulting in a sum value S. A processor determines a single combined modular correction term D for a binary sum of all operand values based on leading bits of the sum value S. A processor performs a modular addition of S and D resulting in a modular sum of said plurality of said first operand values.

Abstract: Embodiments detect a memory constraint in response to a live kernel update (LKU) operation being planned; acquire memory from an external entity through a hardware connection in response to detecting the memory constraint; perform the LKU operation using the acquired memory from the external entity; release the acquired memory; and output a success message which indicates that the LKU operation has been successfully completed in response to the acquired memory being released.

Abstract: A computer-implemented method, program product, and systems for managing distributed processes is provided. Embodiments of the present invention define additional system parameters to a kernel using a union structure. Embodiments of the present invention can then process information manager kernel using a cluster model using a formed distance between one node of the plurality and another node of the plurality. Embodiments of the present invention can then control an interval for information collection and sorting frequency for each node of the plurality of nodes in the system.

Abstract: A trace assist unit operable with a plurality of processor cores is described. The trace assist unit comprises a plurality of physical buffers, and loading circuitry and unloading circuitry that are communicatively coupled with the plurality of physical buffers. The loading circuitry receives trace events from various ones of the plurality of processor cores, each of the trace events having a respective category from a plurality of predefined categories. The loading circuitry writes the trace events to respective ones of the plurality of physical buffers that are assigned to the respective categories of the plurality of predefined categories. The loading circuitry transmits, responsive to one or more predefined conditions, an unload signal to the unloading circuitry to unload contents of a selected physical buffer of the plurality of physical buffers to an external memory.

Abstract: Embodiments of the invention include a stacked device having a first epitaxial region and a second epitaxial region vertically displaced from the first epitaxial region. The first epitaxial region comprising an asymmetric profile with a horizontal protrusion.

Abstract: One embodiment of the invention provides a method for federated learning (FL) comprising training a machine learning (ML) model collaboratively by initiating a round of FL across data parties. Each data party is allocated tokens to utilize during the training. The method further comprises maintaining, for each data party, a corresponding data usage profile indicative of an amount of data the data party consumed during the training and a corresponding participation profile indicative of an amount of data the data party provided during the training. The method further comprises selectively allocating new tokens to the data parties based on each participation profile maintained, selectively allocating additional new tokens to the data parties based on each data usage profile maintained, and reimbursing one or more tokens utilized during the training to the data parties based on one or more measurements of accuracy of the ML model.

Abstract: A system and method for dynamically visualize and modify user authority and security resource edge based on access behaviors in a cloud computing environment. The method provides a new access control perspective to ensure the credibility of database object usage patterns, build clearer and explicit security data consumption mode for data usage requests, and guide tenants to have a more comprehensive understanding and concise use. Processing components and method are configured to gather data and analytics including collecting data resource and database object usage tagging both resources and users with security level by machine learning and building an access model for each role. The method uses graph modeling with created, trained access patterns for users protected resources and produces in-memory implicit sub-groups for users with similar access requirements and limits their resource accesses according to their history behaviors.

Abstract: An approach for assisting users with physical activities by using one or more exoskeleton suits is disclosed. The approach includes, receiving a plurality of data from a user and devices; analyzing one or more problems associated with the plurality of data; determining one or more solutions for the one or more problems based on the analysis; identifying an optimal solution from the one or more solutions; instructing the one or more exoskeleton suits with the optimal solution, wherein the user wears the one or more exoskeleton suits; and performing the instructions, by the exoskeleton suits, based on the optimal solution.

Abstract: A semiconductor structure includes a first plurality of backside power rail interconnects located within a first cell height region of a substrate. A second plurality of backside power rail interconnects are located within a second cell height region of the substrate. A first isolation region is located between the first cell height region of the substrate and the second cell height region of the substrate. The first isolation region electrically separates the first cell height region and the second cell height region. A second isolation region is located between adjacent power rail interconnects of the first plurality of backside power rail interconnects and between adjacent power rail interconnects of the second plurality of backside power rail interconnects. The second isolation region electrically separates the adjacent power rail interconnects.

Abstract: Resource optimized forwarding of classified requests to different instances of a microservice is provided. The process includes obtaining classified requests to a microservice. The classified requests are classified based on different microservice resource consumption types. Further, the method includes forwarding the classified requests to instances of the microservice with different assigned resource configurations. The forwarding is based on the different microservice resource consumption types of the classified requests.

Abstract: An execution-step sequence for a first graphical user interface can be determined. A second graphical user interface can include: first panel in which the first graphical user interface is presented; a second panel including a first user selectable component via which a user navigates forward in the execution-step sequence for the first graphical user interface and a second user selectable component via which the user navigates backward in the execution-step sequence for the first graphical user interface; and a third panel configured to represent steps of the execution-step sequence that are executed. Responsive to the user selecting the first user selectable component, an execution-step in the execution-step sequence can be executed, a result of the executing the execution-step can be depicted by redrawing the first graphical user interface, and component status data indicating the execution-step in the execution-step sequence that was executed can be added to the third panel.

Abstract: Concepts are proposed for augmenting a plurality of climate impact and hazard models. In particular, the plurality of climate impact and hazard models are modified/altered based on a user-specified requirement (i.e., a specified geographical location, temporal location or hazard type) and at least one intelligent (i.e., AI-enabled) workflow. Each modified climate impact and hazard model is then analyzed/processed using a machine learning model to identify model parameters which may improve the climate impact and machine learning model. An operation is then identified and executed based on the intelligent workflow, and model parameter. In this way, speed, efficiency and accuracy of climate impact and hazard models may be improved at scale.

Abstract: Systems and methods for modulating conductance of a plurality of unit cells are described. A data center coupled to a low Earth orbit (LEO) satellite can determine an activation window of an application. The data center can map the orbital movements of the LEO satellite to the activation window of the application. The data center can allocate a resource group to the application. The data center can schedule a server in the data center to run the application according to a result of the mapping and the allocation of the resource group.

Abstract: An embodiment for executing a driver's expectations in view of contextual data and driver expectation data to improve driver safety. The embodiment may automatically establish a connection between an autonomous vehicle and a brain computer interface (BCI) device worn by a driver of the autonomous vehicle. The embodiment may automatically utilize sensors and the BCI device to gather contextual data, automatically process the contextual data, and conduct predictive analysis of correlated driver behavior. The embodiment may automatically detect a traffic context and generate a predicted driver expectation for the traffic context. The embodiment may, in response to detecting that the predicted driver expectation for the traffic context is safe, automatically execute the predicted driver expectation. The embodiment may, in response to detecting that the predicted driver expectation is unsafe, automatically execute a safety remedial action.

Abstract: In an approach to improve improving impact prevention in both manual and autonomous vehicles embodiments of the present invention generate a knowledge corpus based on data collected, from one or more internet of thing (IoT) sensors, associated with one or more predetermined risk factors of one or more predefined risks over time. Further, embodiments receive from the one or more IoT sensors input data to observe a vertical space above a vehicle and identify an object in a predetermined area based on the received input data and the knowledge corpus. Additionally, utilize the knowledge corpus and the input data to generate a risk score for the identified object and responsive to the risk score being above a predetermined threshold, embodiments issue a command for the vehicle to continue to drive or to progress to a predetermined safe zone.

Abstract: A method, computer system, and a computer program product for brake detection is provided. The present invention may include determining a first geometric ratio associated with a rear portion of a first vehicle, from a first image of the first vehicle. The present invention may also include determining a second geometric ratio associated with the rear portion of the first vehicle, from a second image of the first vehicle. The present invention may further include, in response to determining that the second geometric ratio is less than the first geometric ratio, detecting a braking action of the first vehicle.

Abstract: A memory array includes a word lines intersecting bit line-complementary bit line pairs at a plurality of cell locations. Magnetic tunnel junction cells are located at each location, and each cell is electrically connected to a corresponding bit line and selectively interconnected to a corresponding one of the complementary bit lines under control of a corresponding one of the word lines. Each cell includes a substrate; a seed layer overlying the substrate; a nitride layer, overlying the seed layer, and having a thickness greater than 5 Angstroms; a templating layer, outward of the nitride layer, including a binary alloy having an alternating layer lattice structure, and having a thickness greater than 50 Angstroms; a magnetic layer overlying the templating layer, including a Heusler compound and exhibiting PMA; a tunnel barrier outward of the magnetic layer; and a magnetic layer outward of the tunnel barrier.

Abstract: A microelectronic structure including a stacked device region, where stacked device region is comprised of a plurality of top devices and a plurality of bottom devices. Each of the plurality of top devices includes at least one top source/drain. Each of the plurality of bottom devices includes at least one bottom source/drain. A gate cut region located adjacent to the stacked region and an interconnect located in the gate cut region. The interconnect is connected to at least two different devices located in the stacked device region.

Abstract: Embodiments of present invention provide a method of forming a MRAM structure. The method includes patterning a bottom electrode layer and a first ferromagnetic layer on top of the bottom electrode layer; depositing a dielectric layer, the dielectric layer covering the bottom electrode layer and the first ferromagnetic layer; creating an opening in the dielectric layer, the opening exposing a portion of the first ferromagnetic layer; forming a tunnel barrier layer inside the opening; forming a second ferromagnetic layer on top of the tunnel barrier layer; patterning the tunnel barrier layer and the second ferromagnetic layer; and forming a top electrode layer on top of the second ferromagnetic layer. Structures formed thereby are also provided.

Abstract: A chip is manufactured using a method for forming a back-end-of-line (BEOL) layer on an IC chip surface comprises providing a first layer on top of a substrate layer of the IC chip, the first layer comprising a bottom portion of a metallic fill region having a first width as seen in a vertical cross-section of the IC chip. The method further provides a second layer on top of the first layer. The second layer comprises a middle portion of the metallic fill region having a second width that is wider than the bottom portion of the metallic fill region. The method provides a third layer on top of the second layer. The third layer comprises a top portion of the metallic fill region having a third width as seen in the vertical cross-section of the IC chip that is narrower than the middle portion of the metallic fill region.

Abstract: A computer-implemented method for optimizing query performance in a data virtualization system. A processor set generates a number of query collections comprising query statements. The processor set determines network transfer times for data between the data virtualization system and a number of data sources connected with the data virtualization system, and an execution time for executing each query collection. The processor set identifies a number of factors associated with query performance for the data virtualization system. The processor set trains a machine learning model for estimating execution times for the queries based on the number of factors, the network transfer times, and the execution times. The processor set evaluates network conditions and performances of the number of data sources using the machine learning model. The processor set adjusts delegation of the queries to be executed remotely using the number of data sources based on the evaluation.

Abstract: Systems and techniques that facilitate ASR capitalization and punctuation are provided. For example, one or more embodiments described herein can comprise a system, which can comprise a memory that can store computer executable components. The system can also comprise a processor, operably coupled to the memory that can execute the computer executable components stored in memory. The components can comprise a segmenter training component that trains a segmenter machine learning submodel with a set of speech transcriptions and segment boundary information of the speech transcriptions; the segmenter machine learning submodel that generates text segment boundary information of a text comprising a plurality of sentences; and a capitalizer training component that transforms the text into a plurality of segments based on the text segment boundary information and trains a capitalizer machine learning submodel of the text segmenter and capitalizer machine learning model with the plurality of segments.

Abstract: Disclosed embodiments provide techniques for creating a smaller version of an original document. An ontology is defined, and a document type for an original document is determined, based on the ontology. Multiple key-value pairs (KVPs) are extracted from the original document based on the document type. Pages in the original document that include one or more KVPs are identified, and the smaller, condensed version includes the identified pages from the original document.

Abstract: A piezoelectric and thermoelectric solar panel assembly includes a solar panel, a cold mirror, and a support panel. The piezoelectric and thermoelectric solar panel assembly also includes piezoelectric generator assembly disposed between the cold mirror and the support panel. The piezoelectric and thermoelectric solar panel assembly further includes a thermoelectric power generation module beneath the cold mirror.

Abstract: An architecture for demultiplexing volumes in a storage device includes a virtual machine. A guest operating system is in the virtual machine. The architecture further includes a storage pool of virtual storage blocks and a hypervisor. The hypervisor is configured to organize the storge pool of virtual storage blocks into a plurality of virtual storage partition tables. Device addresses are assigned to the plurality of virtual storage partition tables. The plurality of virtual storage partition tables and the devices addresses are organized into a virtual storage device. The device addresses are presented to the guest operating system in the virtual machine.

Abstract: Noisy data parsed from raw data can be received. The noisy data indicates first transaction events determined to be noise in the raw data. Using the noisy data, a first detection model can be trained to assign anomaly event indicators to second transaction events. The first detection model can receive an anomaly record. The anomaly record can indicate at least a portion of anomalous transaction events identified in runtime data. The first detection mode can assign the anomaly event indicators to the anomalous transaction events. The anomaly event indicators can indicate levels of severity of the anomalous transaction events identified in the runtime data.

Abstract: In an approach for policy security shifting left of infrastructure as code compliance, a processor trains a neural network model to classify a code per policy and provide a policy vector score for the code associated with one or more policies. A processor enables the neural network model to scan and score a new code during a continuous integration and continuous deployment pipeline. A processor outputs a scanned score of the new code to a user. A processor retrains the neural network model by capturing a continuous integration and continuous deployment change and run-time compliance posture that occurs as a response by the user.

Abstract: Embodiments are provided for providing enhanced reasoning in a computing system by a processor. All first-order logic formulas may be converted into real-valued logic formulas. A probabilistic inference is executed using the real-valued logic formulas and one or more probability intervals associated with an atomic formulae in a knowledge base to provide an interval conditional probability indicating that a first predicate condition is true based one or more alternative predicates being true.

Abstract: A semiconductor IC device includes a conductive through device connection. The connection may be located within a double diffusion break (DDB) region that separates active regions. The connection may include a faux S/D region between a frontside contact and a backside contact. The semiconductor IC device may further include a first and/or second diffusion break isolation rail. The connection may be between the first and second diffusion break isolation rails. The connection location within the DDB region may resultantly increase packing densities of the semiconductor IC device. Further, the connection may reduce routing complexities and resistance through the semiconductor IC device, which may improve semiconductor IC device performance. Further, the connection may utilize mirrored structure instances (e.g., frontside contact, backside contact, faux S/D region, or the like) as that are used by microdevices (e.g., transistors, or the like) within the active regions, which may decrease fabrication complexities.

Abstract: A semiconductor interconnect structure and formation thereof. The semiconductor interconnect structure includes a first high aspect ratio metal line. The first high aspect ratio metal line includes a first low aspect ratio line segment and a second low aspect ratio line segment.

Abstract: Provided are techniques for enhancing silent features with adversarial networks for improved model versions. Input features are obtained. Hidden features are identified. Quantum feature importance scoring is performed to assign an importance score to each of the hidden features. Silent features are identified as the hidden features with the importance score below a first threshold. Important features are identified as the input features and as the hidden features with the importance score above a second threshold. A silent feature model is built using the silent features. An important feature model is built using the important features. An ensemble model is built with the silent feature model and the important feature model. The ensemble model is used to generate one or more predictions and one or more prescriptions.

Abstract: A computer-implemented method, according to one embodiment, includes using a trained semantic consistency model to determine a semantic consistency of contents of two screenshots. The trained semantic consistency model bases the semantic consistency on dimensions including: a linguistic comparison of the contents, an image comparison of the contents and a text location comparison of the contents. The method further includes outputting the determined semantic consistency for display on a user device. A computer program product, according to another embodiment, includes a computer readable storage medium having program instructions embodied therewith. The program instructions are readable and/or executable by a computer to cause the computer to perform the foregoing method. A system, according to another embodiment, includes a processor, and logic integrated with the processor, executable by the processor, or integrated with and executable by the processor.

Abstract: Techniques for intelligently managing task automation across multiple tasks and multiple automated assistants are provided. In one aspect, an intelligent multitasking system includes: a context manager configured to use human-centric input data to determine a context of a user; an intent mapper configured to map human communication to automated task intents; and an interruption manager configured to schedule automated tasks for performance by automated assistants based on the context of the user and the automated task intents. The context manager can be hosted on a cloud having MQTT clients (e.g., IoT sensors) and an MQTT broker. The intent mapper can include a human-machine software communication interface that identifies verbal and/or non-verbal communications. A method for intelligent multitasking is also provided.

Abstract: Provided are techniques for replicating data with a data orchestrator. A replication request and a policy for an original dataset to be replicated on one or more storage systems as one or more replicated datasets are received. A storage system of the one or more storage systems is selected. It is determined whether a custom storage system plug-in is associated with the storage system based on a storage system type. In response to determining that the custom storage system plug-in is associated with the storage system, the custom storage system plug-in is executed to create a replicated dataset on the storage system. In response to determining that the custom storage system plug-in is not associated with the storage system, a script is generated based on the policy and the script is executed to create the replicated dataset on the storage system.

Abstract: A processor set is configured to receive a plurality of software modules to be analyzed and select one or more software modules that contain modified code. The processor set may further parse product documentation or code comments related to the selected software modules and use the parsed product documentation or code comments to generate a dependency matrix (DM). The processor set may also tag test cases with an associated directed acyclic graph (DAG) that includes a sequence of modules directly or indirectly impacted by the test cases. By reconciling the DM and the DAG for the test cases, the processor set may generate a consolidated directed acyclic graph (CDAG). The processor set may determine, and perform, an optimized list of test cases, based on the generated CDAG and one or more consolidated dependency matrix (CDM) nodes.

Abstract: A structure is provided that includes a thermal dissipation measurement sensor located on the frontside or the backside of a semiconductor device. The thermal dissipation measurement sensor includes a first probe region and a second probe region that are interconnected by a plurality of metal lines that are arranged in a serpentine pattern and having at least one non-powered metal line located between, and spaced apart from, each of the metal lines.

Abstract: A method, computer system, and a computer program product for published content protection is provided. The present invention may include receiving a content file from a content management system (CMS). The present invention may include extracting a feature from the received content file. The present invention may include transforming, using an adversarial generation algorithm, the received content file into an adversarial content file. The present invention may include returning the adversarial content file to the CMS. The returned adversarial content file may represent an equivalent of the received content file to a content consumer. The present invention may include preventing an application of the returned adversarial content file in at least one machine learning task based on the adversarial noise included in the returned adversarial content file.

Abstract: A computer implemented method determines resource affinity for an application is provided. A number of processor units determine connections in affinity data for the workloads for an application. The number of processor units assign weights to the connections based on a number of hops for the connections. The weights indicate the affinity to components related to the set of workloads.

Abstract: A method, a structure, and a computer system for dynamic media routing. The exemplary embodiments may include determining a quality of one or more existing network connections corresponding to one or more participants of a virtual conference and transmitting a media stream sample and a corresponding checksum value between the one or more members of the virtual conference using one or more alternative network connections. The exemplary embodiments may further include determining a quality of the one or more alternative network connections based on the media stream sample and the corresponding checksum value, and, based on determining that the quality of the one or more alternative network connections is higher than the quality of the one or more existing network connections, switching the one or more existing network connections to the one or more alternative network connections.

Abstract: A method of forming a transistor structure is provided. The method includes forming on a substrate first and second mandrels for forming two-dimensional (2D) transistor fin elements defining a pitch gap region, depositing and anisotropically etching back the first spacer material to form first and second spacers in and around the first and second mandrels, respectively, conformally depositing and anisotropically etching back second spacer material around the first and second spacers and in the pitch gap region to define space for forming an odd number of one-dimensional (1D) transistor fin elements in the pitch gap region and depositing and anisotropically etching back the first spacer material in the space with enough cycles to fill the space to form a third spacer.

Abstract: A method, computer system, and a computer program product for management and delivery of augmented reality (AR) content is provided. Embodiments may include receiving, by a first smart device, a request from an augmented reality (AR) user device to view a first device AR content specific to the first smart device. Embodiments may also include, retrieving, by the first smart device, the first device AR content from a local data store of the first smart device. Embodiments may further include, transmitting, by the first smart device, the first device AR content from the local data store of the first smart device to the AR user device.

Abstract: In an approach, a processor, for each object: determines a closest oscillator, subscribes to receive, from the object, an audio data stream with an embedded data structure inserted at a rate determined by the closest oscillator, by a user, wherein the user is associated with a device and the avatar, determines a first distance from the object to the avatar, creates the embedded data structure based on the first distance, the embedded data structure comprising a set of audio parameters, streams the audio data stream with the embedded data structure to the device, and processes the audio data stream at the device according to the embedded data structure to determine a processed audio data stream. A processor mixes a set of processed audio data streams from each of the set of objects to determine a resulting audio data stream. A processor plays the resulting audio data stream at the device.

Abstract: Generating, using a computer, a digital twin model or simulation for organ transplant or organ repairs where organ data of an organ for a patient is received at a computer. Organ parts are identified based on organ data, and the organ parts include a retained portion of the organ and a replacement portion of the organ. A digital model is generated as a digital twin simulations of the organ parts based on the received organ data. Parameters for the organ parts is determined, which include dimensions of the retained portion of the organ and dimensions of the replacement portion of the organ.

Abstract: In an approach to mixed reality scenario generation for cross industry training, a computer receives data associated with tasks performed by a user. A computer identifies an activity in a first industry associated with the tasks. A computer determines that one or more activities associated with a job in a second industry are similar to the activity in the first industry. A computer generates a mixed reality scenario representing a virtual work environment associated with the second industry using mixed reality. A computer generates the one or more activities associated with the job in the second industry in the mixed reality scenario. A computer displays the virtual work environment associated with the second industry and the one or more activities associated with the job in the second industry to the user in a device for viewing a mixed reality scenario while the user performs the activity in the first industry.