Abstract: A substrate structure is produced by forming a first material layer on a substrate having a recess, removing the first material layer from the portion of the substrate except for the recess using a second material that reacts with the first material, and forming a deposition film from the first material layer using a third material that reacts with the first material. A method of manufacturing a device may include the method of forming a substrate structure.

Abstract: A system includes a time-interleaved device. An equalizer effectively can apply different equalization to different interleaved channels. For convenience, these equalizers will be referred to as multi-channel equalizers. In one aspect, an apparatus includes an interleaved device having M interleaved channels, and a multi-channel equalizer coupled to the interleaved device. The multi-channel equalizer is capable of applying a different equalization to different interleaved channels, thus compensating for channel-dependent impairments.

Abstract: A high-voltage LDMOSFET includes a semiconductor substrate, in which a gate well is formed. A source well and a drain well are formed on either side of the gate well, and include insulating regions within them that do not reach the full depth. An insulating layer is disposed on the substrate, covering the gate well and a portion of the source well and the drain well. A conductive gate is disposed on the insulating layer. Biasing wells are formed adjacent the source well and the drain well. A deep well is formed in the substrate such that it communicates with the biasing wells and the gate well, while extending under the source well and the drain well, such as to avoid them. Biasing contacts at the top of the biasing wells bias the deep well, and therefore also the gate well.

Abstract: Reference frames are selected for use in a frame reference buffer in a video encoding system, so as to provide long term prediction of motion in video sequences. Candidate reference frames are characterized by characteristic vectors. From the candidate reference frames, a number of reference frames are maintained in the frame buffer so as to obtain a specified distribution of the characteristic vectors over a vector space. Frames are selected for use in encoding a current frame based on the characteristic vector of the current frame and the reference frames in the frame buffer.

Abstract: One or more behavior-based features describing an execution of an application on a client are generated. One or more client-specific features are generated, wherein the client-specific features describe aspects of the client. A malware score is generated based on the behavior-based features and the client-specific features. Whether the application is a malware threat is determined based on the malware score and a result of the determination is stored.

Abstract: A system, method, and computer program product verifies web page containers of software gadgets. A request for a software gadget is received by a gadget server from a web browser at a client. The request includes a container web page parameter that identifies a web page. The gadget server determines whether the identified web page is an acceptable web page for containing the requested software gadget. If it is determined to be acceptable, the requested software gadget is prepared by inserting container verification code into the gadget. When the prepared software gadget is provided to the client, the container verification code executes and ensures that the requested software gadget is contained within the identified web page. The container verification code may include code for setting a container web page of the prepared software gadget to be the identified web page and for preventing unnecessary page reloads.

Abstract: Systems and methods for performing a trick mode of video streams transmitted over a network without increasing the amount of data transmitted over the network. A video server transrates a source video stream to a target video stream by removing pictures from the source video stream. The target video stream has a reduced number of pictures compared to the source video stream. Therefore, when the target video stream is played on a display device, the target video stream has a playback speed faster than the playback speed of the source video stream.

Abstract: A switch controller is disclosed that adaptively controls the operating frequency of a switching power converter in order to improve one-time load response and repetitive dynamic load responses. During a transition from a high load to low load condition, the switch controller clamps the operating frequency of the switching power converter at an intermediate frequency for a period of time before allowing the operating frequency to return to a frequency associated with the low load condition. The clamped frequency is higher than the frequency associated with the low load condition thereby allowing improved response to a subsequent load change to a high load condition. Thus, the system improves dynamic load response without compromising no-load power consumption.

Abstract: Reading data from an eBook is used to generate time-based information. Timing reports are received from a plurality of clients. The timing reports describe reading time intervals for reading portions of an eBook on the clients. Timing reports from different clients describe reading time intervals for different overlapping portions of the eBook. The overlapping portions are analyzed to generate timing information. The timing information describes reading speeds for granular portions of the book that are smaller than the portions described in the reading time intervals. The timing information is stored. Timing information is provided to a client enabling the client to provide advanced features. The features include estimating a user's current reading position and providing an estimate of the time required by the user to finished reading a portion of the eBook.

Abstract: A power amplifier controller circuit controls a power amplifier based upon an amplitude correction signal indicating the amplitude difference between the amplitude of the input signal and an attenuated amplitude of the output signal. The power amplifier controller circuit comprises an amplitude control loop and a phase control loop. The amplitude control loop adjusts the supply voltage to the power amplifier based upon the amplitude correction signal. The amplitude correction signal may also be split into two or more signals with different frequency ranges and provided respectively to different types of power supplies with different efficiencies to generate the adjusted supply voltage to the power amplifier. The phase control loop adjusts the phase of the input signal based upon a phase error signal indicating a phase difference between phases of the input signal and the output signal to reduce phase distortion generated by the power amplifier.

Abstract: Timing behaviors associated with constraints of circuits are compared to identify mismatches between circuit configurations. Aggregate sets of timing constraints associated with timing nodes are determined for timing paths between start points and end points. Precedence rules are applied to aggregate sets of timing constraints by applying precedence rules to interacting timing constraints. Aggregate sets of constraints for corresponding timing nodes are matched to determine if timing constraint mismatches exist between circuits. If aggregate timing constraints associated with start point, end point pairs are found to match, reconvergent points between the start point and end points are analyzed to see if aggregate constraints of timing nodes connected to reconvergent/divergent points match if timing exception matches are involved. Graph traversal algorithms allow efficient computation of aggregate timing constraints for timing nodes.

Abstract: Provisioning resources in public cloud infrastructure to perform at least part of electronic design automation (EDA) tasks on the public cloud infrastructure. The provisioning of resources is handled by a cloud provisioning system that is generally operated and maintained by an EDA tool developer using a provisioning credential. After the resources are provisioned, the cloud provisioning system places user key on the provisioned resources. Once the user key is placed on the provisioned resources, the cloud provisioning system has only limited access or no access to the provisioned resources. Instead, a user client device takes over the control of the provisioned resources by using a user's access credential. The provisioning credential is retained by the EDA tool developer and is not released to the user. Similarly, the access credential is retained by the user and not released to the EDA tool developer.

Abstract: A non-volatile memory (NVM) circuit is provided, that includes at least a first and second NVM sub-array. The first sub-array is built from first memory cells. The second NVM sub-array is built from second memory cells that are constructed differently from the first memory cells. The NVM sub-arrays share a support circuit. In some embodiments the sub-arrays can be constructed, so that they exhibit different characteristics tailored to their intended use. For example one sub-array might be tailored for data retention, while the next sub-array for programming endurance, still another for write disturb immunity.

Abstract: A method for forming a thin film using radicals generated by plasma may include generating radicals of a reactant precursor using plasma; forming a first thin film on a substrate by exposing the substrate to a mixture of the radicals of the reactant precursor and a source precursor; exposing the substrate to the source precursor; and forming a second thin film on the substrate by exposing the substrate to the mixture of the radicals of the reactant precursor and the source precursor. Since the substrate is exposed to the source precursor between the formation of the first thin film and the formation of the second thin film, the rate of deposition may be improved.

Abstract: Dehydration of routes enables transmitting a description of a route requiring much less space than full specification of the route. A series of “breadcrumbs” and hints are used for dehydration. A breadcrumb includes coordinates of a point, a heading at which the route enters the breadcrumb, and a heading at which the route leaves the breadcrumb. A dehydration module places a breadcrumb at the location marking the beginning of the route, and having a leaving heading identifying the link in the original route. The node at the end of each link in the original route is examined. If the link leaving the node is the most parallel link to the link entering the node, nothing is added to the dehydrated route. If not, a breadcrumb is added to the dehydrated route, specifying the coordinates of the point, the entering heading of the breadcrumb and the leaving heading of the breadcrumb.

Abstract: An optical circulator couples optical fibers of dissimilar modes. In one embodiment, an optical circulator couples a single mode first fiber to a multimode second fiber, which is used as an optical input to a telescope. The multimode fiber does not significantly degrade the mode structure of the light form the single mode fiber. In the reverse direction, light received by the telescope is coupled into the multimode second fiber, which the circulator couples to a multimode third fiber.

Abstract: A pen based computing system concurrently captures handwriting gestures and records audio using binaural recording. A binaural headset communicatively coupled to the smart pen device uses at least two microphones. A left microphone is placed in or near the left ear and the right microphone is placed in or near the right ear, each facing outward. Speakers are integrated into a shared housing with the microphones facing inward towards the ear canal to play back the audio recordings. By recording audio with microphones placed close to the ears, the system provides realistic sounding playback and allows users to more easily differentiate between multiple sources of audio.

Abstract: An object recognition system performs a number of rounds of dimensionality reduction and consistency learning on visual content items such as videos and still images, resulting in a set of feature vectors that accurately predict the presence of a visual object represented by a given object name within an visual content item. The feature vectors are stored in association with the object name which they represent and with an indication of the number of rounds of dimensionality reduction and consistency learning that produced them. The feature vectors and the indication can be used for various purposes, such as quickly determining a visual content item containing a visual representation of a given object name.

Abstract: Semantically meaningful segments of an image of a document, such as tables of contents, page numbers, footnotes, and the like, are identified. These segments form a model of the document image, which may then be rendered differently for different client devices. The rendering may be based on a display parameter provided by a client device, such as a display resolution of the client device, or a requested display format.