Abstract: A processor can include various processing pipelines that perform different data processing operations, with different pipelines having dedicated logic and memory circuits. A power management circuit can determine when to supply power to various pipelines, including the logic and memory circuits of the various pipelines, depending on a current operating mode of the processor. When a memory circuit transitions to a lower power state such as a sleep state, data can be saved to a different memory circuit that is not transitioning to a lower power state, and when the memory circuit is powered up again, the data can be restored from the different memory circuit.

Abstract: A method for adaptive mapping for data compression includes determining an input/output (I/O) request pattern, dynamically switching between a segment mapping mode and a flat hash table mapping mode based on the determined I/O request pattern, updating a shared mapping table for the segment mapping mode and the flat hash table mapping mode, and adjusting an entry of the mapping table based on the determined I/O request pattern and a status of the entry.

Abstract: Methods, systems, and apparatus, including computer-readable media, are described for processing an input image using a convolutional neural network (CNN). The CNN includes a sequence of layer blocks. Each of a first subset of the layer blocks in the sequence is configured to perform operations that include: i) receiving an input feature map for the layer block, ii) generating an expanded feature map from the input feature map using a group convolution, and iii) generating a reduced feature map from the expanded feature map. The input feature map is an h w feature map with c1 channels. The expanded feature map is an h w feature map with c2 channels, whereas the reduced feature map is an h w feature map with c1 channels. C2 is greater than c1. An output feature map is generated for the layer block from the reduced feature map.

Abstract: According to an aspect, there is provided a method performed by a network function consumer, NFC. The method comprises sending (701) a request (401) to a network function producer, NFP, wherein the request (401) is for context information for a first wireless device, and wherein the request (401) comprises one or more filter values identifying a subset of context information for the first wireless device required by the NFC. The subset of context information for the first wireless device required by the NF consumer is related to specific Protocol Data Unit, PDU, sessions.

Abstract: A method for accessing blocks of a solid state drive is described. A starting position is received, where the starting position identifies a first block of a contiguous block region within a namespace of the solid state drive. The contiguous block region is to be accessed according to a single input/output operation and the namespace comprises two dimensions of logical address space with respective indices for indexing blocks within a corresponding dimension of the logical address space. A first dimensional identifier that identifies a size of the contiguous block region in a first dimension of the namespace is received. A second dimensional identifier that identifies a size of the contiguous block region in a second dimension of the namespace is received. Blocks of the contiguous block region are accessed in response to the single input/output operation according to the starting position, the first dimensional identifier, and the second dimensional identifier.

Abstract: The invention relates to a traffic flow machine-learning modeling system and method applied to vehicles. The system includes: a target fusion module configured to perform target fusion on radar measurement data and camera measurement data, and output target attribute information; a lane line model module configured to output an original lane line model based on the camera measurement data; a target selection module configured to determine a position of a lane where each target is located based on the target attribute information output by the target fusion module and the original lane line model output by the lane line model module, and output the target; and a traffic flow calculation module configured to model the vehicle position by using a clustering algorithm based on the output of the target fusion module, the output of the target selection module, and the output of the lane line model module, and output a traffic flow-based lane line model.

Abstract: A solar photovoltaic module comprises a generator body and a support body, and can be repaired in place after installation on a roof by removing and replacing the solar panel laminate from the generator body of the solar module. The support body of the solar module comprises an elongated rail defining an upper tabletop onto which a fastener may be affixed and a lower tabletop to receive the generator body. The solar panel laminate of the generator body is retained by a frame comprising a horizontal front frame element having a hook to engage a support body of another solar module.

Abstract: The invention relates to a method for lane-changing prediction of a target vehicle, the method including: receiving a velocity and a position of the target vehicle; respectively obtaining, based on the velocity and the position of the target vehicle, a first lane-changing probability and a second lane-changing probability of the target vehicle by using a first machine learning model and a second machine learning model; and determining a possibility of lane changing of the target vehicle based on the first lane-changing probability and the second lane-changing probability, the first machine learning model and the second machine learning model being pre-trained and being different from each other. The invention further relates to a device for lane-changing prediction of a target vehicle, a computer storage medium, and a vehicle.

Abstract: An optical communication module, a device and a system are provided. The optical communication module includes: an electrical communication unit, an exchanging communication mainboard, a first optical transceiving unit and a second optical transceiving unit. The first optical transceiving unit and the second optical transceiving unit respectively receive an optical signal from and/or send an optical signal to a first device and a second device and complete conversion from or to the optical signal, and by the first optical transceiving unit and the second optical transceiving unit, devices connected with the electrical communication unit, the first device and the second device form a network that has a function of a self-healing ring network.

Abstract: A method for determining driving parameters includes: acquiring driving data of a driver; extracting driving features of the driver based on the driving data, where the driving features include first operation frequency of a first component of a vehicle and second operation frequency of a second component of the vehicle; determining, based on the driving features, a first probability that the driver has a first driving style and a second probability that the driver has a second driving style; and determining the driving parameters based on the first probability and the second probability, where the driving parameters include at least longitudinal acceleration and longitudinal deceleration. The method can dynamically adjust automated driving parameters to meet the driving style of the driver, thereby effectively improving user experience.

Abstract: The present disclosure describes techniques of metadata management for transparent block level compression. A first area may be created in a backend solid state drive. The first area may comprise a plurality of entries. The plurality of entries may be indexed by addresses of a plurality of blocks of uncompressed data. Each of the plurality of entries comprises a first part configured to store metadata and a second part configured to store compressed data. Each of the plurality blocks of uncompressed data may be compressed individually to generate a plurality of compressed blocks. Metadata and at least a portion of compressed data associated with each of the plurality of compressed blocks may be stored in one of the plurality of entries based on an address of a corresponding block of uncompressed data. A second area may be created in the backend solid state drive for storing the rest of the compressed data.

Abstract: Systems and methods for space allocation for block device compression are provided. In particular, a computing device may receive an allocation request to write the compressed data, select a range list adequate for serving the allocation request from a plurality of range list, dequeue a range entry from the selected range list to allocate free space for the compressed data, and allocate the free space corresponding to the range entry to the compressed data to serve the allocation request.

Abstract: Methods and systems for adaptive mapping for data compression on a storage device is provided. The method includes determining a data request pattern of a workload, determining whether to use at least one of a segment mapping mode or a hash mapping mode for mapping the workload, dividing a space on the storage device into a plurality of defrag units for storing data, and assigning the plurality of defrag units as being at least one of a segment defrag unit or a hash defrag unit. The method also includes when the data request pattern is for the segment mapping mode, storing the data on at least one of the plurality of defrag units assigned as the segment defrag unit, and when the data request pattern is for the hash mapping mode, storing the data on at least one of the plurality of defrag units assigned as the hash defrag unit.

Abstract: A system and method are described to efficiently allocate memory space with low latency overhead by allocating blocks of non-volatile memory on a storage device according to a tree data structure comprising a plurality of counter sets, each counter set including one or a plurality of counters indicating numbers of unallocated blocks of memory space within the non-volatile memory.

Abstract: Embodiments herein describe methods, devices, and systems for rupture detection and end-of-life monitoring of dynamic gas lock (DGL) membranes and pupil facet mirrors in lithographic apparatuses. A method for detecting rupture of a dynamic gas lock membrane in a lithographic apparatus includes illuminating the dynamic gas lock membrane with a measurement beam using a radiation source, in which the dynamic gas lock membrane is arranged between a wafer and projection optics of the lithography apparatus, and determining whether any radiation from the measurement beam is reflected from the dynamic gas lock membrane by using reflection collection optics, in which the reflection collection optics are arranged above the dynamic gas lock membrane. A rupture in the dynamic gas lock membrane is detected if no radiation is reflected from the dynamic gas lock membrane. If radiation is reflected from the dynamic gas lock membrane, the dynamic gas lock membrane is not ruptured.

Abstract: Methods, systems, and apparatus, including computer-readable media, are described for processing an input image using integrated circuit that implements a convolutional neural network with a group convolution layer. The processing includes determining a mapping of partitions along a channel dimension of an input feature map to multiply accumulate cells (MACs) in a computational unit of the circuit and applying a group convolution to the input feature map. Applying the group convolution includes, for each partition: providing weights for the group convolution layer to a subset of MACs based on the mapping; providing, via an input bus of the circuit, an input of the feature map to each MAC in the subset; and computing, at each MAC in the subset, a product using the input and a weight for the group convolution layer. An output feature map is generated for the group convolution layer based on an accumulation of products.

Abstract: Methods, systems, and devices for wireless communications are described. A user equipment (UE) may identify a throttling indicator for a shared resource associated with a modem at the UE. The CE may determine a configuration of the modem based on a configuration received from a Radio Resource Control layer of the UE. Based on the modem configuration and the throttling indicator, the UE may identify a throttling level to apply to the modem. The UE may adjust the modem configuration for communicating with a base station based on the throttling level in order to reduce resource usage below a usage threshold. By reducing usage of the shared resource by the modem, the UE may have a greater availability of the shared resource to meet the needs of other components of the UE. This may provide an improved user experience at the UE, among other benefits.

Abstract: Described are examples for storing data on a storage device, including storing, in a live write stream cache, one or more logical blocks (LBs) corresponding to a data segment, writing, for each LB in the data segment, a cache element of a cache entry that points to the LB in the live write stream cache, where the cache entry includes multiple cache elements corresponding to the multiple LBs of the data segment, writing, for the cache entry, a table entry in a mapping table that points to the cache entry, and when a storage policy is triggered for the cache entry, writing the multiple LBs, pointed to by each cache element of the cache entry, to a stream for storing as contiguous LBs on the storage device, and updating the table entry to point to a physical address of a first LB of the contiguous LBs on the storage device.