Abstract: Recommending database access patterns for transactions in microservices is provided. A recommendation of a particular database access pattern to improve access by transactions to data in fields of tables in a database corresponding to a microservice is generated based on analysis of field access graphs for the tables accessed by the transactions. The recommendation of the particular database access pattern is output. The particular database access pattern is implemented to improve access by the transactions to the data in the fields of the tables in the database corresponding to the microservice to improve performance of the microservice based on user input and read view implementation guidance.

Abstract: An interconnect structure is provided. The interconnect structure includes first conducting lines and second conducting lines. The first conducting lines are formed of a first metallic material and include at least one individual first conducting line in contact with a first corresponding substrate conducting line. The second conducting lines are formed of a second metallic material and include at least one individual second conducting line between neighboring first conducting lines and in contact with a second corresponding substrate conducting line. The at least one individual second conducting line is separated from each of the neighboring first conducting lines by controlled distances.

Abstract: A burn-in resilient integrated circuit is provided. The burn-in resilient integrated circuit includes an inverter chain and a plurality of inverter circuits on the inverter chain. The burn-in resilient integrated circuit also includes a loop providing an electrical connection from an output of the inverter chain to an input of the inverter chain. The loop is selectable in accordance with a burn-in operation.

Abstract: Embodiments include predictive antenna diode insertion. Aspects of the invention include obtaining a design of a macro, the design including an internal pin disposed on a first layer of the macro. Aspects of the invention also include determining a length of a wire needed to connect the internal pin to a furthest edge of the macro for each of two layers above the layer the internal pin. Aspects of the invention further include adding, to the design of the macro, an antenna diode to the internal pin based on the determination that an area of the wire needed exceeds a threshold value, wherein the area of the wire is based on the length and a width of the wire.

Abstract: In an approach to enhancing server virtualization through host sessions by re-screening a remote desktop session of a target virtual machine. Embodiments, identify, on a computer system, an open remote desktop window corresponding to the remote desktop session, and determine, by the computer system, whether one or more second remote desktop windows are embedded in the open remote desktop window. Further, embodiments identify, from the one or more second remote desktop windows, a target remote desktop window associated with the target virtual machine and generate an overlay on the open remote desktop window based on the identification of the target remote desktop window associated with the target virtual machine. Additionally, embodiments capture activity on a remote desktop screen associated with the generated overlay, and executes the captured activity on the generated overlay.

Abstract: Methods, systems, and computer program products for providing the status of model extraction in the presence of colluding users are provided herein. A computer-implemented method includes generating, for each of multiple users, a summary of user input to a machine learning model; comparing the generated summaries to boundaries of multiple feature classes within an input space of the machine learning model; computing correspondence metrics based at least in part on the comparisons; identifying, based at least in part on the computed metrics, one or more of the multiple users as candidates for extracting portions of the machine learning model in an adversarial manner; and generating and outputting an alert, based on the identified users, to an entity related to the machine learning model.

Abstract: Examples described herein provide a computer-implemented method that includes training a machine learning model. The model is trained by generating a set of training queries using at least one of a query workload and relationships between tables in a database, building a query graph for each of the set of training queries, computing, for each training query of the set of training queries, a selectivity based at least in part on the query graph, and building, based at least in part on the set of training queries, an initial join result distribution as a collection of query graphs.

Abstract: An embodiment includes building a hierarchical data structure using a hybrid hierarchical clustering process. The hybrid hierarchical clustering process comprises one or more iterations of a level-building process. An embodiment of the level-building process comprises building, by a classical processor, a first parent level of a current uppermost level of the hierarchical data structure by clustering classically-encoded clusters of the current uppermost level. The embodiment of the level-building process also comprises identifying, by a quantum processor, a set of candidate clustering options for clustering quantum-encoded clusters of the current uppermost level for a second parent level, including forming each of the set of candidate clustering options in parallel using respective different quantum encoding spaces. The embodiment of the level-building process also comprises building, by the classical processor, the second parent level based on a subset of the candidate clustering options.

Abstract: A method, system and apparatus for protecting against out-of-bounds references, including storing an address of a buffer in a general register and storing bounds information (BI) for the buffer in a bounds information register, and when a content of the general register is used as an address in a load or store operation, using a content of the bounds information register to determine if the load or store is out of bounds.

Abstract: According to at least one embodiment, a method, computer system, and computer program product for model-driven extraction of event data by a computational device is provided. The present invention may include transforming event data of an event occurring on a blockchain network using an event schema to comply with a data model of an external data structure; mapping, based on the event schema and the data model, the event data to a logical structure of the external data structure; and loading the transformed event data to the external data structure to modify a data content of the external data structure based on the mapping.

Abstract: A method, computer system, and a computer program product for online posting annotation is provided. The present invention may include identifying a topic in an online posting. The present invention may include determining that a comment to the online posting relates to the online posting. The present invention may include adding at least one detail from the comment to the online posting. The present invention may include tagging the at least one added detail with at least one piece of metadata. The present invention may include validating a viewing access based on the at least one piece of metadata and displaying an amended online posting.

Abstract: Illustrative embodiments provide techniques for fabricating semiconductor structures having bottom isolation and enhanced carrier mobility for both nFET and pFET devices. For example, in one illustrative embodiment, a semiconductor structure includes a semiconductor substrate, a first dielectric layer disposed on the semiconductor substrate, a bottom source/drain region disposed on the first dielectric layer and isolated from the semiconductor substrate by the first dielectric layer, a second dielectric layer disposed on the bottom source/drain region and a top source/drain region disposed on the second dielectric layer and isolated from the bottom source/drain region by the second dielectric layer. The bottom source/drain region comprises a compressive pFET epitaxy and the top source/drain region comprises a tensile nFET epitaxy.

Abstract: A computer system controls access to data. A request is received from an entity to access data comprising a primary data object corresponding to a physical item. One or more secondary data objects included in the primary data object are identified, wherein the one or more secondary data objects correspond to physical components of the physical item. Access requirements for the primary data object and the one or more secondary data objects are determined. In response to determining that the access requirements are satisfied by the entity, the entity is granted access to the data comprising the primary data object and the one or more secondary data objects. Embodiments of the present invention further include a method and program product for controlling access to data in substantially the same manner described above.

Abstract: A method, apparatus and computer program product for scheduling placement of containers in association with a set of hosts. The technique utilizes metrics that characterize container-specific risks. A first metric is a host interface risk for a container that quantifies how similar or dissimilar the container is relative to other containers running on a host. Preferably, host interface risk is derived with respect to a system call interface comprising a set of system calls, and the metric is based at least in part on a measure of dissimilarity among system calls. A second metric is a data sensitivity score that quantifies a degree to which sensitive data accesses are associated to the container. Based at least in part on the host interface risk scores and the data sensitivity scores, one or more containers are automatically scheduled for placement on the set of hosts to minimize security risk for the set of hosts.

Abstract: An approach for selecting a transformed prediction model for estimating the uncertainty of prediction associated with machine learning is disclosed. The approach determines a function corresponding to a prediction interval based on one or more datasets and calculates one or more sets of prediction interval parameters associated with the function based on training a set of regression models with the one or more datasets. The approach creates one or more transformed predictions based on the one or more sets of parameters and based on a logical distance and selects a transformed prediction model based on a mean absolute correlation from the one or more transformed predictions. Furthermore, the approach outputs the selected transformed prediction model.

Abstract: Provided is a tape drive reel including a damping structure for reducing position error signals in a tape during operation. The tape drive reel includes a hub including a first flange and a second flange. At least one of the first flange and the second flange includes a damping layer and a single stiffener layer positioned thereon. The damping layer is positioned between the single stiffener layer and the at least one of the first flange and the second flange. The single stiffener layer is more rigid than the damping layer.

Abstract: According to one embodiment, a method, computer system, and computer program product for personalizing playback of an audio stream is provided. The present invention may include sectioning the audio stream into one or more content blocks and one or more filler blocks; determining one or more topics associated with the one or more filler blocks; determining a level of complexity associated with the one or more filler blocks; determining a listener's level of interest in and level of comprehension of the one or more topics; based on the level of complexity, the level of interest, and level of comprehension corresponding to the one or more topics, assigning a playback speed to the one or more content blocks associated with the one or more topics; and modifying the one or more content blocks to play at the assigned playback speed.

Abstract: Method and system are provided for efficiently populating a phase diagram for modeling of multiple substances. The method may include defining an n-way phase diagram with data points each being an n-tuple describing the n substance inputs, wherein the n-way phase diagram is defined at a user-configured resolution. The method may select an initial subset of data points and calculate their contribution to the phase diagram. The method may then generate a Bayesian model based on the initial subset of calculated data points and predicting the resultant phase and an associated uncertainty of all the uncalculated data points in the defined phase diagram. The method may select a sample subset of the data points using maximum entropy sampling and calculating a resultant phase for each of the selected data points, and incorporate the calculated phases into the Bayesian model.

Abstract: A method and system perform single phase and multi-phase contour refinement of lesions. The method includes receiving a three dimensional input mask; receiving input slices from the medical images including a lesion; cropping the input slices with the input mask; performing lesion contour refinement for the cropped input slices and the input mask to obtain a predicted mask; and storing the predicted mask that includes 3D lesion contour refinement. A multiphase method includes deforming the 3D input mask from the reference phase to a target phase or warping the input slices from the target phase to the reference phase before contour refinement. The warped images generate an output mask in the reference phase coordinate system that is then deformed to the target phase coordinate system for display.

Abstract: A method of via formation including forming a sacrificial mask over a conductive layer, forming a plurality of pillars in the sacrificial mask and the conductive layer, wherein each pillar of the plurality of pillars includes a sacrificial cap and a first conductive via, depositing a spacer between the plurality of pillars, masking at least one of the sacrificial caps, removing at least one of the sacrificial caps to create openings, forming second conductive vias in the openings, and depositing a dielectric coplanar to a top surface of the second conductive vias.

Abstract: A method is presented for forming a vertical transport field effect transistor (VTFET). The method includes forming a plurality of fins over a substrate, depositing a sacrificial material adjacent the plurality of fins, forming self-aligned spacers adjacent the plurality of fins, removing the sacrificial material to define openings under the self-aligned spacers, filling the openings with bottom spacers, depositing an interlayer dielectric (ILD) after patterning, laterally etching the substrate such that bottom surfaces of the plurality of fins are exposed, the lateral etching defining cavities within the substrate, and filling the cavities with an epitaxial material such that epitaxial regions are defined each having a symmetric tapered shape under a twin-fin structure. The single fin device can be formed through additional patterning and bottom epi under the single fin device that has an asymmetric tapered shape.

Abstract: A semiconductor structure includes a substrate comprising a semiconductor material, and a fin on the substrate. The fin includes a first portion formed from the semiconductor material and a second portion including a channel region. The first portion has a first thickness and the second portion has a second thickness greater than the first thickness. A spacer is disposed on sides of the first portion of the fin.

Abstract: An approach for representing both positive and negative weights in neuromorphic computing is disclosed. The approach leverages a double gate FeFET (ferroelectric field effect transistor) device. The device leverages a double-gate FeFET with four terminals (two separate gates and source and drain) and ferroelectric gate dielectric. The device may have a junction-less channel. A synaptic weight is programmed by biasing one of the two gates. The store weight is sensed via a current flow from source to drain. A pre-defined bias is applied to the other gate during the sensing, such that a reference current is subtracted from the drain current. The net current for sensing is current from the synaptic devices subtracted by the pre-defined reference current.

Abstract: A method for identifying hard-coded strings in source code is disclosed. In one embodiment, such a method parses source code and associated localization resource files to identify hard-coded strings and their associated context. The method provides a confidence score for each hard-coded string that indicates whether the hard-coded string is translatable or non-translatable. Based on the confidence score for each hard-coded string, the method transforms each hard-coded string into a single equivalence word. The method then prepares training data by tagging the hard-coded strings in the source code and associated localization resource files as one of translatable and non-translatable. The method then trains a parts-of-speech (POS) tagging model using the training data. At runtime, the method fetches potential hard-coded strings and tags each hard-coded string as one of translatable and non-translatable using the POS tagging model. A corresponding system and computer program product are also disclosed.

Abstract: An embodiment includes creating an action item record corresponding to an action item of an action plan record that is responsive to a service request. The action item record comprises a service requirement of the action item. The embodiment executes a querying process that searches vendor records for candidate vendors associated with the service requirement and returns a set of candidate vendors. The embodiment updates the action item record with the set of candidate vendors and determines an optimal vendor team based at least in part on reputation data and cost data associated with each of the candidate vendors. The embodiment updates the action plan record to include the optimal vendor team, which triggers creation of a vendor team dispatch request.

Abstract: Various embodiments are provided for configuring optimization problems from one or more sources in a computing environment by a processor. A knowledge graph may be generated from a knowledge domain and one or more data sources. One or more graph pattens may be applied to match one or more entities in the knowledge graph with one or more atomic optimization templates. An optimization problem configured from the one or more atomic optimization templates and a plurality of data may be executed.

Abstract: A method for minimizing data transfer and storage utilization on cloud object storage systems is disclosed. In one embodiment, such a method replicates a file from a production system to a cloud object storage system. The method determines whether a number of hard links associated with the file is greater than one. In the event the number is greater than one, the method creates, on the cloud object storage system, a special object for the file and associates the file with the special object. Upon creating a hard link on the production system in association with the file, the method replicates the hard link from the production system to the cloud object storage system without replicating data associated with the file. A metadata reference to the special object is added to the hard link on the cloud object storage system. A corresponding system and computer program product are also disclosed.

Abstract: A generator network of a variational autoencoder can be trained to approximate a simulator and generate a first result. The simulator is associated with input data, based on which the simulator outputs output data. A training data set for the generator network can include the simulator's input data and output data. Based on the simulator's output data and the first result of the generator network, an inference network of the variational autoencoder can be trained to generate a second result. The second result of the trained inference network inverts the first result of the generator and approximates the simulator's input data. The trained inference network can function as an inverted simulator.

Abstract: Provided is a system and method which prioritizes data replication packets between a private cloud and a public cloud which provides a backup for the private cloud. In one example, the method may include receiving a request from a software application to write data to a storage location of a private cloud that hosts the software application, identifying storage attributes of the storage location of the private cloud, generating a replication request for replicating the data over a network to a public cloud, embedding a priority tag into the replication request based on the identified storage attributes of the storage location of the private cloud, and transmitting the tagged replication request over the network from the private cloud to the public cloud based on a bandwidth assigned to the embedded priority tag.

Abstract: A computer-implemented method, computer program product, and computer system for accessing content on a web page. The method accesses a displayed web page in a web browser of a user computing device and receives a search term for content within the web page. The method accesses contextual information held locally on the user computing device and scans the web page to identify one or more portions of content of the web page relevant to the search term and the contextual information. The method provides a display on the user computing device with links to the relevant web page content.

Abstract: A method, a computer program product, and a system for pre-pooling virtual machine components prior to application startup. The method includes reading a virtual machine pool manifest by a virtual machine pool manager. The virtual machine pool manifest includes an initial number of virtual machines to launch, how many virtual machines to allocate resources to, and an amount of resources to allocate based on a virtual machine resource definition. The method also includes launching a plurality of base virtual machines into a virtual machine pool and based on the initial number provided by the virtual machine pool manifest. The base virtual machines lack initial application assignments. The method further includes allocating initial resources to a portion of the base virtual machines based on a virtual machine resource definition in the virtual machine pool manifest. The method includes loading core program packages into the portion of the base virtual machines.

Abstract: In at least some embodiments, a store-type operation is received and buffered within a store queue entry of a store queue associated with a cache memory of a processor core capable of executing multiple simultaneous hardware threads. A thread identifier indicating a particular hardware thread among the multiple hardware threads that issued the store-type operation is recorded. An indication of whether the store queue entry is a most recently allocated store queue entry for buffering store-type operations of the hardware thread is also maintained. While the indication indicates the store queue entry is a most recently allocated store queue entry for buffering store-type operations of the particular hardware thread, the store queue extends a duration of a store gathering window applicable to the store queue entry. For example, the duration may be extended by decreasing a rate at which the store gathering window applicable to the store queue entry ends.

Abstract: Techniques and apparatus for analyzing visual content using a visual transformer are described. An example technique includes generating a first set of tokens based on a visual content item. Each token in the first set of tokens is associated with a regional feature from a different region of a plurality of regions of the visual content item. A second set of tokens is generated based on the visual content item. Each token in the second set of tokens is associated with a local feature from one of the plurality of regions of the visual content item. At least one feature map is generated for the visual content item, based on analyzing the first set of tokens and the second set of tokens separately using a hierarchical vision transformer. At least one vision task is performed based on the at least one feature map.

Abstract: A semiconductor device is provided. The semiconductor device includes a bottom epitaxial layer, a gate stack formed over the bottom epitaxial layer, the gate stack including a work function metal (WFM) layer, a channel fin formed on the bottom epitaxial layer, a first interlayer dielectric (ILD) layer formed in a gate landing area over the gate stack, a second ILD layer formed in an area other than the gate landing area, and a WFM encapsulation layer formed between the first ILD layer and the second ILD layer, and formed on sidewalls of the gate stack.

Abstract: A system, method, and apparatus for concealing searches for information stored on public networks, includes splitting of a search query, transforming the query into one or more related queries, and searching each of related queries separately.

Abstract: A semiconductor device is provided. The semiconductor device includes: a first conductive electrode; a first dielectric stack structure provided on the first conductive electrode; a second conductive electrode provided on the first dielectric stack structure; a second dielectric stack structure provided on the second conductive electrode; and a third conductive electrode provided on the first dielectric stack structure, wherein each of the first dielectric stack structure and the second dielectric stack structure include a first dielectric layer comprising a first material; a second ferroelectric dielectric layer comprising a second material and provided on the first dielectric layer, and a third dielectric layer comprising a third material and provided on the second ferroelectric dielectric layer.

Abstract: Techniques regarding integrated purification-detection devices for detecting one or more biomarkers are provided. For example, one or more embodiments described herein are directed to an apparatus, comprising a housing and a microfluidic chip contained within the housing. The microfluidic chip comprises a separation unit that separates, using one or more nano deterministic lateral displacement (nanoDLD) arrays, target biological entities having a defined size range from other biological entities included in a biological fluid sample. The microfluidic chip further comprises a detection unit that facilitates detecting presence of one or more biomarkers associated with the target biological entities using one or more detection molecules or macromolecules that chemically reacts with the one or more biomarkers.

Abstract: A user requests to join a meeting is detected. The meeting includes a meeting audio stream of one or more participant audio streams that include participant timestamps that correspond to when one or more other users are in the meeting. The user is prompted for an authentication credential based on the detecting the request to join the meeting. A participant profile of the user is determined based the authentication credential. The user is authorized access to the meeting and a first timestamp is saved. A first audio stream of the user is recorded. The user is identified as having left the meeting and a second timestamp is saved. A transcript of the meeting audio stream is generated based on the first audio stream and the one or more participant audio streams. The first timestamp, the second timestamp, and the meeting are associated with the participant profile.

Abstract: A computer program product, device, system, and method are provided for decommissioning a tape cartridge internally within a tape library. In one embodiment automated tape cartridge decommissioning operations internal within a tape library, include receiving selection of a tape cartridge for decommissioning in a tape library and, using a tape library robotic accessor, moving a selected tape cartridge to a decommission workstation within the tape library. The selected tape cartridge is mounted to the decommission workstation and a fusing element of the decommission workstation is activated to fuse together layers of magnetic tape of the selected tape cartridge mounted to the decommission workstation, so that the selected tape cartridge is decommissioned within the tape library to have a decommissioned status.

Abstract: A spin-orbit torque magnetoresistive random-access memory device formed by fabricating a spin-Hall-effect (SHE) layer above and in electrical contact with a transistor, forming a spin-orbit-torque (SOT) magnetoresistive random access memory (MRAM) cell stack disposed above and in electrical contact with the SHE rail, wherein the SOT-MRAM cell stack comprises a free layer, a tunnel junction layer, and a reference layer, forming a cylindrical diode structure above and in electrical contact with the SOT-MRAM cell stack, forming a write line disposed in electrical contact with the SHE rail, and forming a read line disposed above and adjacent to an outer cylindrical electrode of the diode structure.

Abstract: Interconnect structures including super vias are formed during back-end-of-line processing using sacrificial placeholders to protect the bottom portions of the super vias while upper portions of the super vias are formed. The sacrificial placeholders are removed and replaced by metal conductors that fill the bottom and upper portions of the super vias.

Abstract: A field effect device is provided. The field effect device includes an active gate structure, a gate contact within the active gate structure, wherein the gate contact is the same height as the active gate structure, and a gate cut dielectric on opposite sides of the gate contact and active gate structure.

Abstract: Classifying and filtering data by categorizing a portion of a data segment according to the content of the portion, correlating each portion with a user content preference dictionary, assigning a content weight to each portion according to the correlation, and masking content portions having a weight below a threshold value.

Abstract: A method of manufacturing a semiconductor device is provided. The method includes forming a first trench partially through a first substrate from a first side of the first substrate. The method also includes widening a bottom portion of the first trench to form a lateral footing area of the first trench. The method includes forming a first metallization in the first trench; forming a second trench through a second substrate from a second side of the second substrate to expose at least a portion of first metallization in an area corresponding to the lateral footing area of the first trench, the second side being opposite to the first side. The method also includes forming a second metallization in the second trench in contact with the first metallization.

Abstract: Methods for cellular network authentication utilizing unlinkable anonymous credentials are disclosed. In embodiments, a method includes: contacting, by a computing device, a mobile device network with a request to connect to the mobile device network; conducting, by the computing device, an interactive credential issuance protocol with an Issuer of the mobile device network to generate an unlinkable anonymous credential; and connecting, by the computing device, to the mobile device network based on a Verifier of the mobile device network verifying the computing device based on the unlinkable anonymous credential.

Abstract: An approach to provide a funnel-shaped spin-transfer torque (STT) magnetoresistive random-access memory (MRAM) device with a dual magnetic tunnel junction. The approach includes providing a metal pillar on a connection to a semiconductor device. The approach includes providing a first reference layer on the metal pillar and on a portion of a first interlayer dielectric adjacent to the metal pillar. The approach includes providing a first tunnel barrier on the first reference layer and a free layer on the first tunnel barrier layer. The approach includes providing a second tunnel barrier on the free layer and a second reference layer on the second tunnel barrier of the semiconductor structure of the funnel-shaped spin-transfer torque MRAM device.

Abstract: An optical connector device is provided. The optical connector device includes a semiconductor package including a receptacle and a lid. The optical connector device also included an adapter attached to the lid of the semiconductor package, and a connector removably attached to the adapter. The adapter includes a convex part adapted to fit into an adapter opening of the lid, an adapter recess adapted to accommodate at least a portion of the connector, and a first retainer in the adapter recess to removably attach the connector to the adapter at a predetermined position. The connector includes an optical fiber array corresponding to the receptacle and extending in a vertical direction with respect to a plane of the semiconductor package, a second retainer used in conjunction with the first retainer, and a biasing member to bias a portion of the connector toward the semiconductor package.

Abstract: Method, apparatus, and computer program product are provided for generating synthetic bathymetry using one or more fractal techniques. In some embodiments, real terrain data generated from sensor-based measurement of terrain associated with but outside a water body are received. This data may include, for example, real terrain data generated from sensor-based measurement (e.g., LiDAR) of at least a portion of an area proximate the water body. One or more fractal dimensions is/are extracted from the real terrain data, and synthetic bathymetry of the water body is generated based on the one or more fractal dimensions using one or more fractal terrain generation techniques. In some embodiments, one or more simulated geological processes is/are applied to the synthetic bathymetry. For example, a simulated sedimentation process and/or a simulated erosion process may be applied to soften the synthetic bathymetry.

Abstract: Artificial neural networks (ANNs) are a distributed computing model in which computation is accomplished with many simple processing units, called neurons, with data embodied by the connections between neurons, called synapses, and by the strength of these connections, the synaptic weights. An attractive implementation of ANNs uses the conductance of non-volatile memory (NVM) elements to record the synaptic weight, with the important multiply—accumulate step performed in place, at the data. In this application, the non-idealities in the response of the NVM such as nonlinearity, saturation, stochasticity and asymmetry in response to programming pulses lead to reduced network performance compared to an ideal network implementation.

Abstract: Techniques that facilitate model support in deep learning are provided. In one example, a system includes a graphics processing unit and a central processing unit memory. The graphics processing unit processes data to train a deep neural network. The central processing unit memory stores a portion of the data to train the deep neural network. The graphics processing unit provides, during a forward pass process of the deep neural network that traverses through a set of layers for the deep neural network from a first layer of the set of layers to a last layer of the set of layers that provides a set of outputs for the deep neural network, input data for a layer from the set of layers for the deep neural network to the central processing unit memory.