Abstract: Systems, apparatuses and methods may provide for technology that detects a low battery condition in a computing system including an integrated graphics processor and a discrete graphics processor, wherein the low battery condition is detected during a pre-boot stage of the computing system. The technology may also disable a root port associated with the discrete graphics processor in response to the low battery condition, conduct an initialization of an integrated display while the root port is disabled, and enable the root port in response to a successful negotiation of increased power by a verified read write code of an embedded controller of the computing system.

Abstract: An on-chip integrated circuit, a data processing device and a method are provided. The on-chip integrated circuit includes: a processor circuit and an accelerator circuit. The processor circuit includes a processor and a data storage area, the processor is connected to the data storage area through a first bus in the processor circuit. The accelerator circuit includes an accelerator and a second bus, the accelerator is connected to the second bus, and the second bus is bridged with the first bus corresponding to the data storage area, to perform data interaction between the accelerator and the data storage area, which can reduce the congestion on a bus of the processor and improve the quality of service of the application.

Abstract: Embodiments described herein include software, firmware, and hardware logic that provides techniques to perform arithmetic on sparse data via a systolic processing unit. One embodiment provides for data aware sparsity via compressed bitstreams. One embodiment provides for block sparse dot product instructions. One embodiment provides for a depth-wise adapter for a systolic array.

Abstract: A method implemented by a computing device comprises determining, by an original thread executing at a computing device, a command used to render a frame of a graphics application, the command being a call to a graphics interface, constructing, by the original thread executing at the computing device, a command stream based on the command, the command stream comprising a plurality of commands used to render the frame, and executing, by a command stream thread executing at the computing device, the command stream to render the frame of the graphics application.

Abstract: Fabric Attached Memory (FAM) provides a pool of memory that can be accessed by one or more processors, such as a graphics processing unit(s) (GPU)(s), over a network fabric. In one instance, a technique is disclosed for using imperfect processors as memory controllers to allow memory, which is local to the imperfect processors, to be accessed by other processors as fabric attached memory. In another instance, memory address compaction is used within the fabric elements to fully utilize the available memory space.

Abstract: An electronic device and a method of operating an electronic device are provided.

Abstract: Embodiments described herein provide techniques to disaggregate an architecture of a system on a chip integrated circuit into multiple distinct chiplets that can be packaged onto a common chassis. In one embodiment, a graphics processing unit or parallel processor is composed from diverse silicon chiplets that are separately manufactured. A chiplet is an at least partially and distinctly packaged integrated circuit that includes distinct units of logic that can be assembled with other chiplets into a larger package. A diverse set of chiplets with different IP core logic can be assembled into a single device.

Abstract: Embodiments described herein provide techniques to disaggregate an architecture of a system on a chip integrated circuit into multiple distinct chiplets that can be packaged onto a common chassis. In one embodiment, a graphics processing unit or parallel processor is composed from diverse silicon chiplets that are separately manufactured. A chiplet is an at least partially and distinctly packaged integrated circuit that includes distinct units of logic that can be assembled with other chiplets into a larger package. A diverse set of chiplets with different IP core logic can be assembled into a single device.

Abstract: A data processing system comprising a plurality of processing units. Each processing unit comprises a set of plural functional units and an internal communications network that routes communications between the functional units in a particular sequence order of the functional units. Each processing unit is connected to at least one other processing unit via a communications bridge that has at least two connections, a first connection that routes communications between a first pair of network nodes of the pair of processing units, and a separate, second connection that routes communications between a second, different pair of network nodes of the pair of processing units. Each connected pair of network nodes comprises network nodes having different positions in the internal communications network sequence order of the network nodes and/or network nodes associated with functional units of different types.

Abstract: An ‘open control network’ is described, wherein the control plane functions within the Radio Access Network (such as eNodeB and gNodeB) and Core Network (such as MME, AMF and SMF) provide an interface towards the operator and 3rd party control applications. Applications are allowed to securely register to signaling protocols within the control plane, specifically to the RAN or the Core Network control functions to view, intercept and intervene certain types of control messages or procedures. Innovative applications can be developed to view and modify control plane behavior utilizing both traditional methods as well as upcoming Machine Learning and Artificial Intelligence algorithms to provide services that are not part of standard operator offerings.

Abstract: A parallel processor system for machine learning includes an arithmetic unit (ALU) array including several ALUs and a controller to provide instructions for the ALUs. The system further includes a direct-access memory (DMA) block containing multiple DMA engines to access an external memory to retrieve data. An input-stream buffer decouples the DMA block from the ALU array and provides aligning and reordering of the retrieved data. The DMA engines operate in parallel and include rasterization logic capable of performing a three-dimensional (3-D) rasterization.

Abstract: Systems, apparatuses, and methods related to a memory array data structure for posit operations are described. Universal number (unum) bit strings, such as posit bit string operands and posit bit strings representing results of arithmetic and/or logical operations performed using the posit bit string operands may be stored in a memory array. Circuitry deployed in a memory device may access the memory array to retrieve the unum bit string operands and/or the results of the arithmetic and/or logical operations performed using the unum bit string operands from the memory array. For instance, an arithmetic operation and/or a logical operation may be performed using a first unum bit string stored in the memory array and a second unum bit string stored in the memory array. The result of the arithmetic operation and/or the logical operation may be stored in the memory array and subsequently retrieved.

Abstract: An arithmetic device and an electronic device having small power consumption is provided. An arithmetic device and an electronic device capable of high-speed operation is provided. An arithmetic device and an electronic device capable of suppressing heat generation is provided. The arithmetic device includes a first arithmetic portion and a second arithmetic portion. The first arithmetic portion includes a first CPU core and a second CPU core. The second arithmetic portion includes a first GPU core and a second GPU core. The CPU cores each have a power gating function and each include a first data retention circuit electrically connected to a flip-flop. The first GPU core includes a second data retention circuit capable of retaining an analog value and reading out the analog value as digital data of two or more bits. The second GPU core includes a third data retention circuit capable of retaining a digital value and reading out the digital value as digital data of one bit.

Abstract: A display driving integrated circuit includes a timing controller, a first source driver including a first inverting input, a first non-inverting input, and a first output, a second source driver including a second inverting input, a second non-inverting input, and a second output, and a switching circuit connected with the display panel through a first and second pads. Under control of the timing controller, the switching circuit performs one of a first switching operation of connecting the first inverting input and the first output with the first pad, connecting the second inverting input and the second output with the second pad, and applying first and second decoding voltages to the non-inverting inputs, respectively; and a second switching operation of applying a sensing reference voltage to the non-inverting inputs, and connecting the output terminals with an output node, and connecting the inverting inputs with one pad.

Abstract: An electronic circuit includes a converter and a controller. The converter outputs a first voltage for a first cluster and a second voltage for a second cluster. When a first power to be provided to the first cluster based on the first voltage is lower than a first available power of the first cluster and a second power to be provided to the second cluster based on the second voltage is higher than a second available power of the second cluster, the controller outputs a first interrupt signal such that a level of the second voltage is adjusted based on a sum of the first power and the second power and a first threshold value determined based on the first available power and the second available power.

Abstract: An erasable magnetoresistive random-access memory (MRAM) structure and a method of making the same includes an MRAM cell disposed between bit line and word line circuit elements, and a vertical-cavity surface-emitting laser (VCSEL) element disposed above the MRAM cell. A laser output of the VCSEL is directed toward the MRAM cell.

Abstract: A method, system and computer program product are disclosed for using a Field Programmable Gate Array (FPGA) to simulate operations of a device under test (DUT). The DUT includes a device memory having a number of input ports, and the FPGA is associated with a target memory having a second number of input ports, the second number being less than the first number. In one embodiment, a given set of inputs is applied to the device memory at a frequency Fd and in a defined cycle of time, and the given set of inputs is applied to the target memory at a frequency Ft. Ft is greater than Fd and cycle accuracy is maintained between the device memory and the target memory. In an embodiment, a cycle accurate model of the DUT memory is created by separating the DUT memory interface protocol from the target memory storage array.

Abstract: Briefly, in accordance with one or more embodiments, a processor receives an incoming data stream that includes alpha channel data, and a memory stores an application programming interface (API). The API is to route the alpha channel data to a fixed point blending unit to perform one or more blending operations using fixed point representation of the alpha channel data. The API is further to route the incoming data stream to a floating point blending unit to perform operations involving floating point representation of the incoming data.

Abstract: A round-robin system having a suspend mechanism. The system may solve a synchronization issue relative to wireless communication with several hosts. With round-robin, a gateway device may permit user communication with one host at a time. The suspend mechanism, during a round-robin sequence, may enable a user with a request via an internet cloud, or the like, to interrupt the sequence and go to one specific host.

Abstract: Virtual reality systems and methods are described. For example, one embodiment of an apparatus comprises: a communications interface to provide frame data of a virtual reality scene to a head mounted display (HMD); at least one performance monitor coupled to at least one component of the apparatus the at least one performance monitor to monitor performance of the at least one component and to send an alert based on the performance of the at least one component; a processor to process the frame data; a controller to receive the alert based on the performance of the at least one component and to offload processing of the frame data from the processor to the HMD for processing; and a display to show the rendered view of the scene.

Abstract: A process of updating data for a block of an array of data elements stored in an allocated memory region for the block comprises reading in data for a first group of the data elements, updating the data for the first group, and then writing back the updated data to memory. The process can avoid overwriting data for a second group of the data elements that is yet to be read in from the memory region by writing back the updated data for the first group starting at a selected memory address, for example other than the start address of the memory region. The data for the second group of data elements can then be read in and updated, and the updated data can be written back to memory. The process can reduce the amount of memory bandwidth and local cache that needs to be used.

Abstract: Embodiments of a system and method for enhanced graphics rendering performance in a hybrid computer system are generally described herein. In some embodiments, a graphical element in a frame, application, or web page, which is to be presented to a user via a web browser, is rendered either by a first processor or a second processor based on indications of whether the first or the second processor is equipped or configured to provide faster rendering. A rendering engine may utilize either processor based on historical or anticipated rendering performance, and may dynamically switch between the hardware decoder and general purpose processor to achieve rendering time performance improvement. Switches between processors may be limited to a fixed number switches or switching frequency.

Abstract: A host system includes a host device, a host buffer memory, and storage device. The host device includes a plurality of cores. The host buffer memory is configured to store a first command queue and a first map table, wherein each of the first command queue and the first map table corresponds to a first core of the cores. The storage device is configured to perform an input/output virtualization operation using the first core as a virtual core. The storage device uses the first command queue and the first map table during the input/output virtualization operation using the first core.

Abstract: System and method for training and performing operations (e.g., read and write operations) on a double buffered memory topology. In some embodiments, eight DIMMs are coupled to a single channel. The training and operations schemes are configured with timing and signaling to allow training and operations with the double buffered memory topology. In some embodiments, the double buffered memory topology includes one or more buffers on a system board (e.g., motherboard).

Abstract: A drive circuit applicable to a display device includes a first signal path and a second signal path. The first signal path, configured to receive and transmit image data, includes a compression unit configured to perform a compression procedure on the image data to generate compression data; a storage unit configured to store the compression data; and a de-compression unit configured to perform a de-compression procedure on the compression data to recover the image data. The second signal path is configured to transmit the image data to the storage unit so as to bypass the compression unit, and transmit the image data received from the storage unit to a display unit so as to bypass the de-compression unit when the image data is not transmitted by the first signal path. The received image data is passed through the first signal path or the second signal path depending upon its characteristics.

Abstract: A tile coalescer within a graphics processing pipeline coalesces coverage data into tiles. The coverage data indicates, for a set of XY positions, whether a graphics primitive covers those XY positions. The tile indicates, for a larger set of XY positions, whether one or more graphics primitives cover those XY positions. The tile coalescer includes coverage data in the tile only once for each XY position, thereby allowing the API ordering of the graphics primitives covering each XY position to be preserved. The tile is then distributed to a set of streaming multiprocessors for shading and blending operations. The different streaming multiprocessors execute thread groups to process the tile. In doing so, those thread groups may perform read-modify-write operations with data stored in memory. Each such thread group is scheduled to execute via atomic operations, and according to the API order of the associated graphics primitives.

Abstract: A network device may include a first forwarding board, a second forwarding board and an interface board. The interface board may include a control apparatus and a network interface chip. The control apparatus may form a first upstream packet flow which is sent to the first forwarding board via a first I/O bus and a second upstream packet flow which is sent to the second forwarding board via a second I/O bus using data packets received through the network interface chip from the exterior of the network device.

Abstract: A parallel data processing method based on multiple graphic processing units (GPUs) is provided, including: creating, in a central processing unit (CPU), a plurality of worker threads for controlling a plurality of worker groups respectively, the worker groups including one or more GPUs; binding each worker thread to a corresponding GPU; loading a plurality of batches of training data from a nonvolatile memory to GPU video memories in the plurality of worker groups; and controlling the plurality of GPUs to perform data processing in parallel through the worker threads. The method can enhance efficiency of multi-GPU parallel data processing. In addition, a parallel data processing apparatus is further provided.

Abstract: A layer selection module for a graphics display component, and method therefor. The layer selection module is arranged to identify a set M of active layers to be blended for a pixel, configure a display controller to generate composite pixel data for the pixel based on a subset N of up to n layers from the set M, determine whether a number m of active layers in the set M exceeds n, and output an indication of which active layers within the set M were excluded from the subset N, if it is determined that the number m of layers in the set M exceeds n.

Abstract: A system that incorporates teachings of the present disclosure may include, for example, a communication device having a controller to determine a directive associated with an incoming communication session initiated by another communication device, providing an option that directs the media device to modify the media content being presented by a media device, providing a first instruction to the media device to modify the media presentation according to the directive responsive to a selection of the option, and providing a second instruction to the media device to resume the media presentation responsive to detecting a termination of the incoming communication session. Other embodiments are disclosed.

Abstract: Methods for preprocessing pixel data using a Direct Memory Access (DMA) engine during a data transfer of the pixel data from a first memory (e.g., a DRAM) to a second memory (e.g., an SRAM) are described. The pixel data may derive from a color camera or a depth camera in which individual pixel values are not a multiple of eight bits. In some cases, the DMA engine may perform a variety of image processing operations on the pixel data prior to the pixel data being written into the second memory. In one embodiment, the DMA engine may be configured to determine whether one or more pixels corresponding with the pixel data may be invalidated or skipped based on a minimum pixel value threshold and a maximum pixel value threshold and to embed pixel skipping information within unused bits of the pixel data.

Abstract: System on chip comprising a general purpose processing element, a graphics processing unit and a display interface, supporting graphics visualization on mobile computing devices and on embedded systems.

Abstract: The present application discloses a processor and a method for executing an instruction on a processor. A specific implementation of the processor includes: a host interaction device, an instruction control device, an off-chip memory, an on-chip cache and an array processing device, wherein the host interaction device is configured to exchange data and instructions with a host connected with the processor, wherein the exchanged data has a granularity of a matrix; the off-chip memory is configured to store a matrix received from the host, on which a matrix operation is to be performed; and the instruction control device is configured to convert an external instruction received from the host to a series of memory access instructions and a series of computing instructions and execute the converted instructions. The implementation can improve the execution efficiency of a deep learning algorithm.

Abstract: A display controller includes a scaler which is configured to receive a frame image, scale up the frame image to generate a high resolution frame image based on a quality of the frame image and information about a display device, and output the frame image or the high resolution frame image to the display device for display. A method of controlling image display includes: receiving a frame image; determining whether to scale up the frame image to generate a high resolution frame image based on a quality of the frame image and information about a display device; scaling up the frame image according to a result of the determining; and outputting the frame image or the high resolution frame image for display at the display device.

Abstract: A tile coalescer within a graphics processing pipeline coalesces coverage data into tiles. The coverage data indicates, for a set of XY positions, whether a graphics primitive covers those XY positions. The tile indicates, for a larger set of XY positions, whether one or more graphics primitives cover those XY positions. The tile coalescer includes coverage data in the tile only once for each XY position, thereby allowing the API ordering of the graphics primitives covering each XY position to be preserved. The tile is then distributed to a set of streaming multiprocessors for shading and blending operations. The different streaming multiprocessors execute thread groups to process the tile. In doing so, those thread groups may perform read-modify-write operations with data stored in memory. Each such thread group is scheduled to execute via atomic operations, and according to the API order of the associated graphics primitives.

Abstract: A tile coalescer within a graphics processing pipeline coalesces coverage data into tiles. The coverage data indicates, for a set of XY positions, whether a graphics primitive covers those XY positions. The tile indicates, for a larger set of XY positions, whether one or more graphics primitives cover those XY positions. The tile coalescer includes coverage data in the tile only once for each XY position, thereby allowing the API ordering of the graphics primitives covering each XY position to be preserved. The tile is then distributed to a set of streaming multiprocessors for shading and blending operations. The different streaming multiprocessors execute thread groups to process the tile. In doing so, those thread groups may perform read-modify-write operations with data stored in memory. Each such thread group is scheduled to execute via atomic operations, and according to the API order of the associated graphics primitives.

Abstract: A tile coalescer within a graphics processing pipeline coalesces coverage data into tiles. The coverage data indicates, for a set of XY positions, whether a graphics primitive covers those XY positions. The tile indicates, for a larger set of XY positions, whether one or more graphics primitives cover those XY positions. The tile coalescer includes coverage data in the tile only once for each XY position, thereby allowing the API ordering of the graphics primitives covering each XY position to be preserved. The tile is then distributed to a set of streaming multiprocessors for shading and blending operations. The different streaming multiprocessors execute thread groups to process the tile. In doing so, those thread groups may perform read-modify-write operations with data stored in memory. Each such thread group is scheduled to execute via atomic operations, and according to the API order of the associated graphics primitives.

Abstract: Convolution neural networks are able to be trained using a GPU and a CPU. To efficiently utilize a device's resources, the HetNet and HybNet approaches have been developed. The HetNet approach separates batches into partitions such that the GPU and CPU process separate batches. The HybNet approach separates the layers of a convolution neural network for the GPU and CPU.

Abstract: A method of updating a file in a solid state drive (SSD) and an SSD configured to update a file in the SSD is disclosed. In one embodiment, the method includes performing one or more writes to a holding file in an auxiliary memory, the one or more writes corresponding to an update for a target file in the auxiliary memory. The method further includes applying the update to the target file in the auxiliary memory when each of the one or more writes has been successfully written to the holding file, and resetting the holding file when less than all of the one or more writes have been successfully written to the holding file. In one embodiment, a flash controller in communication with the auxiliary memory performs the update.

Abstract: A semiconductor apparatus pertaining to one embodiment has: a first processor that operates by a first program and reads pixel data from a storage unit; a second processor that operates by a second program, performs processing to the pixel data, and writes the processed pixel data back to the storage unit; and a buffer circuit that transfers the pixel data from the first processor to the second processor.

Abstract: An apparatus is described. The apparatus includes an image processing unit. The image processing unit includes a network. The image processing unit includes a plurality of stencil processor circuits each comprising an array of execution unit lanes coupled to a two-dimensional shift register array structure to simultaneously process multiple overlapping stencils through execution of program code. The image processing unit includes a plurality of sheet generators respectively coupled between the plurality of stencil processors and the network. The sheet generators are to parse input line groups of image data into input sheets of image data for processing by the stencil processors, and, to form output line groups of image data from output sheets of image data received from the stencil processors.

Abstract: An information handling system, includes a server rack including a first communication device. The server rack has a mounting flange with a plurality of mounting holes. The information handling system also includes a server installed in the server rack and secured to the mounting flange via a first mounting hole of the plurality of mounting holes. The server includes a second communication device, that is coupled to the first communication device via a second mounting hole of the plurality of mounting holes.

Abstract: A display driver integrated circuit (DDI) for driving a display of image data on a display panel, an application processor (AP), a system including the DDI and the AP, and methods of operating the same are provided. The application processor includes: a controller configured to obtain a frequency of a data transmission timing control received from a display driver integrated circuit (DDI), and to generate, based on the obtained frequency, a frequency control signal for adjusting a frequency related to an operating clock signal for the DDI; a transmitter configured to transmit the generated frequency control signal to the DDI; and a frequency calculation circuit including: a detector configured to receive the data transmission timing control signal from the DDI, and a frequency calculator configured to calculate a frequency of the received data transmission timing control signal.

Abstract: A color compensation circuit comprises an acquisition unit for acquiring, from a video signal, gray image of a frame to be displayed and chrominance image of any color; wherein a chrominance value in the video signal, corresponding to at least a portion of pixels within the chrominance image, is absent; a processing unit for smoothly processing the chrominance value in the chrominance image according to the change trend of gray value in the gray image to obtain a chrominance image with color compensated. The present disclosure can solve the problem that compression and decompression of a video signal during transmission will significantly decrease the frame display effect, and thereby helping improve the frame display effect in a transmission scenario with the loss of chrominance value of the video signal.

Abstract: In one embodiment, a client device configured to remotely access a desktop hosted by a server system determines an event related to a user input for a desktop operation directed to the desktop. The client device receives a plurality of updates to a desktop graphical user interface (GUI) from the desktop hosted by the server system. Then, the client device correlates the event to an update in the plurality of updates to the desktop GUI based on a rule in a set of rules correlating events to updates. A metric is monitored for the update and information measured for the metric is stored.

Abstract: A virtual reality system is provided. The virtual reality system includes a host device and a head mounted display apparatus to be worn by a user. The head mounted display apparatus includes a first wireless module, a second wireless module, a multimedia module, a multi-sensing module, and a peripheral hub. The multimedia module receives multimedia content from the host device through the first wireless module. The multi-sensing module obtains sensing information regarding the head mounted display apparatus and the user. The peripheral hub receives communication data from the host device through the second wireless module, and provides the sensing information to the host device through the second wireless module.

Abstract: Certain aspects direct to a computing device, which include a processor, a microcontroller, a random access memory (RAM) having a frame buffer, a video controller configured to read video data from the frame buffer, and a non-volatile memory. The processor, the microcontroller, the RAM, the video controller, and the non-volatile memory are in communication with each other. The non-volatile memory stores an operating system, a media player, and first video data. The processor is configured to load the operating system to the RAM and execute the operating system. While the processor is loading the operating system, the microcontroller is configured to load and execute the media player into the RAM. The media player, when executed by the microcontroller, is configured to read the first video data from the non-volatile memory, and write second video data representing the first video data to the frame buffer.

Abstract: Embodiments related to managing potentially invalid results generated/obtained by a microprocessor during runahead are provided. In one example, a method for operating a microprocessor includes causing the microprocessor to enter runahead upon detection of a runahead event. The example method also includes, during runahead, determining that an operation associated with an instruction referencing a storage location would produce a potentially invalid result based on a value of an architectural poison bit associated with the storage location and performing a different operation in response.

Abstract: An audio processor has a number of ports that are configurable as input or output ports. Each port includes a jack, an input audio circuit and an output audio circuit. A switch is controllable to selectively connect an output of the output audio circuit to the jack when the port is configured as an output port. In one embodiment, the switch is bypassed with resistor and the output of the output audio circuit is coupled through the resistor to the jack when the port is configured as an input port.

Abstract: A semiconductor device includes: first and second memory cell regions disposed adjacent to each other in a first direction, and suitable for sharing a sub-word line driving signal, and a first sub-word line driving unit disposed in a crossing area that is disposed between the first and second memory cell regions in a diagonal direction. The first sub-word line driving unit includes a first sub-word line driver for driving the first memory cell regions, a second sub-word line driver for driving the second memory cell regions, and an interconnection for transmitting the sub-word line driving signal, which extends in the first direction.