Abstract: Systems and methods are provided for optimizing GPU memory allocation for high-performance applications such as deep learning (DL) computing. For example, a DL task is executed using GPU resources (GPU device and GPU memory) to process a DL model having functional layers that are processed in a predefined sequence. A current functional layer of the DL model is invoked and processed using the GPU device. In response to the invoking, a data compression operation is performed to compress data of a previous functional layer of the DL model, and store the compressed data in the GPU memory. Responsive to the invoking, compressed data of a next functional layer of the DL model is accessed from the GPU memory and a data decompression operation is performed to decompress the compressed data for subsequent processing of the next functional layer of the DL model by the GPU device.

Abstract: Systems, apparatuses, and methods for implementing flexible dictionary sharing techniques for caches are disclosed. A set-associative cache includes a dictionary for each data array set. When a cache line is to be allocated in the cache, a cache controller determines to which set a base index of the cache line address maps. Then, a selector unit determines which dictionary of a group of dictionaries stored by those sets neighboring this set would achieve the most compression for the cache line. This dictionary is then selected to compress the cache line. An offset is added to the base index of the cache line to generate a full index in order to map the cache line to the set corresponding to this chosen dictionary. The compressed cache line is stored in this set with the chosen dictionary, and the offset is stored in the corresponding tag array entry.

Abstract: Apparatus and method for efficient graphics processing including ray tracing. For example, one embodiment of a graphics processor comprises: execution hardware logic to execute graphics commands and render images; an interface to couple functional units of the execution hardware logic to a tiled resource; and a tiled resource manager to manage access by the functional units to the tiled resource, a functional unit of the execution hardware logic to generate a request with a hash identifier (ID) to request access to a portion of the tiled resource, wherein the tiled resource manager is to determine whether a portion of the tiled resource identified by the hash ID exists, and if not, to allocate a new portion of the tiled resource and associate the new portion with the hash ID.

Abstract: A method, computer system, and a computer program product for projecting a 3D model defined by x, y, z coordinates onto the surface of a 2D image defined by u, v coordinates is provided. The present invention may include receiving a 3D model having a plurality of polygons, wherein certain edges are marked as seams. The present invention may include receiving input from a user, wherein the input comprises painting one or more parts of the 3D model in different colors, wherein the colors correspond with a weight of the area painted. The present invention may include unwrapping, by a processor, a 2D texture from the 3D model using a projection algorithm. The present invention may include generating a rectangular boundary around each island. The present invention may include scaling each island according to a gradient score.

Abstract: Methods, systems, and storage media for outline character printing on a printing device are disclosed. Exemplary implementations may: allocate cache memory of a printing device for registration of an outline character; register outline character data wherein the metadata portion comprises a description of the outline character and the data portion comprises a scanline table for the outline character; receive a print job data at the printing device; determine the print job data identifies the outline character in the cache memory at the printing device; render the scanline table accessed from the cache memory as a pattern of pixels for the outline character for the print job data received; and print, on the printing device, the pattern of pixels rendered from the scanline table accessed from the cache memory for the print job data received.

Abstract: Provided are a low-power display driving circuit performing internal encoding and decoding and an operating method thereof. The display driving circuit includes a memory configured to store an input bit stream encoded by an encoder and a controller configured to determine a data path through which output frame data in a second frame period passes according to whether internal encoding is successful in a first frame period, wherein, when the internal encoding is successful, the controller performs internal encoding in the second frame period, stores a generated internal bit stream in the memory, allows the internal bit stream to pass through a low-power path to generate the output frame data, and when the internal encoding fails, the controller generates the output frame data by allowing the input bit stream to pass through a normal path in the second frame period, changes an encoding setting of an internal encoder, and repeats the internal encoding.

Abstract: An apparatus to facilitate compute compression is disclosed. The apparatus includes a graphics processing unit including mapping logic to map a first block of integer pixel data to a compression block and compression logic to compress the compression block.

Abstract: A method and apparatus are provided for compressing vertex parameter data in a 3D computer graphic system, where the vertex parameter data is a data block relating to a plurality of vertices used for rendering an image. The data relating to each vertex includes multiple byte data relating to at least one parameter. The parameters include X, Y and Z coordinates and further coordinates for texturing and shading. The multiple byte data is divided into individual bytes and bytes with corresponding byte positions relating to each vertex are grouped together to form a plurality of byte blocks.

Abstract: An apparatus includes a graphics driver circuit and a graphics engine circuit. The graphics engine circuit is configured to determine graphics data to be output to a display and to render the graphics data to a buffer. The graphics driver circuit is configured to output the buffer to the display. The graphics engine circuit is further configured to, while the graphics driver circuit is outputting the first buffer to the display, encode the first graphics data into another buffer, and to signal the graphics driver circuit to output the other buffer to the display.

Abstract: Methods for lossy and lossless pre-processing of image data. In one embodiment, a method for lossy pre-processing image data, where the method may include, at a computing device: receiving the image data, where the image data includes a model having a mesh, the mesh includes vertices defining a surface, the vertices including attribute vectors, and the attribute vectors including values. The method also including quantizing the values of the attribute vectors to produce modified values, where a precision of the modified values is determined based on a largest power determined using a largest exponent of the values, encoding pairs of the modified values into two corresponding units of information. The method also including, for each pair of the pairs of the modified values, serially storing the two corresponding units of information as a data stream into a buffer, and compressing the data stream in the buffer.

Abstract: According to certain embodiments, reducing data signal bandwidth comprises receiving a multi-resolution image having a plurality of concentric regions that each have a different level of resolution, where the regions closer to the center having greater levels of resolution. Generating a representative image comprising first rasterizing the multi-resolution image into a pixel representation of pixel data with an array of pixel values. A ring of pixels for each region results from discarding duplicate pixels from each region such that only the interior pixels from each region are preserved. In a first circular direction, duplicate pixels along each ring are discarded, the preserved unique pixels resulting in ring fragments. The ring fragments are then moved towards the center so that there are no gaps between the ring fragments for each level of resolution.

Abstract: Disclosed herein is a point cloud data transmission method including encoding the point cloud data, encapsulating the point cloud data, and transmitting point cloud data. Disclosed herein is a point cloud data reception device including a receiver configured to receive the point cloud data, a decapsulator configured to decapsulate the point cloud data, and a decoder configured to decode the point cloud data.

Abstract: In one example embodiment, an information processing apparatus, for an observed image associated with an observation target object (e.g., a section of biological tissue), associates and stores position information and observation magnification information. In this embodiment, the information processing apparatus causes a display device to: (i) display an image associated with the observation target object; (ii) indicate the first positional information of the first observed image; and (iii) indicate the first observation magnification information of the first observed image.

Abstract: A visual rendering apparatus such as a telescope, microscope or attached tablet/led displays a magnified subject using the mapped rendering medium, in which the rendering medium includes at least one of actual visual transmissions of the subject and stored, high resolution images of the magnified subject. In an educational context, equipment for displaying true magnified images of, for example, celestial bodies or molecular structures can be beyond reach. Augmented reality provided by supplementing the true, rendered magnified subject with stored images corresponding to successive, higher magnification levels provides effective visualization with common educational tools, avoiding the need for extravagant scientific equipment.

Abstract: A device implementing the subject high bit-depth graphics compression may include at least one processor configured to receive pixel data for a pixel block, obtain endpoints of a first bit length based on the pixel data in the pixel block, quantize the endpoints to a second bit length smaller than the first bit length, select the quantized endpoints for pixel values in the pixel block, determine a weight for each pixel of the pixel block in each of a plurality of planes corresponding to the endpoints selected for the pixel block, and generate a compressed data block representative of the pixel block based at least on the endpoints for the pixel block and the weight for each pixel of the pixel block in each of the plurality of planes corresponding to the endpoints. A method and computer program product implementing the subject high bit-depth graphics compression is also provided.

Abstract: Methods and apparatus relating to techniques for provision of low power foveated rendering to save power on GPU (Graphics Processing Unit) and/or display are described. In various embodiment, brightness/contrast, color intensity, and/or compression ratio applied to pixels in a fovea region are different than those applied in regions surrounding the fovea region. Other embodiments are also disclosed and claimed.

Abstract: A method and an apparatus for converting image data of a super-pixel array into image data a sub-pixel array are provided. A plurality of nodes in a separate activation area on a sensing circuit are activated to obtain a plurality of super-pixels and to generate image data of a super-pixel array including the plurality of super-pixels. Also, one or more nodes in an integrated activation area are separately activated to obtain values of sub-pixels in a first area of a sub-pixel array, and based on the super-pixel array and the values of the sub-pixels in the first area, the image data of the super-pixel array is converted into image data of the sub-pixel array.

Abstract: A display device includes a display panel including pixels, a luminance controller that divides the display panel into blocks based on coordinate information, calculates a block reference current based on a block current sensed in each of the blocks when reference images are sequentially displayed on the blocks, calculates a target current based on the block reference current and a block load of each of the blocks based on input image data, and calculates a scaling factor based on the target current and a sensing current sensed in each of the blocks when an input image corresponding to the input image data is displayed on the display panel, and a data driver that generates a data voltage corresponding to the input image data and supplies the data voltage to the pixels by adjusting a voltage level of the data voltage based on the scaling factor.

Abstract: A mobile device such as a mobile phone is used to image a nearby user for eye tracking at a closer range than would be provided by cameras on a display device such as a TV and send the eye tracking data to a server streaming video being presented on the display. The eye tracking data is used by the server to identify a region of interest (ROI) in the video, which is more highly compressed than regions outside the ROI to facilitate efficient compression of the video game image while preserving the quality of the ROI in the image where the user is looking.

Abstract: A method for adapting display data forming an image for display on a display screen to a viewer involves monitoring, over time, light that may affect an eye of the viewer of the image, analysing (S82) information regarding the monitored light based on one or more predetermined parameters (which may be based on a predetermined model of reactions of a human eye to light or to changes in light), adjusting (S83) at least one display value (preferably not a luminance value) of at least some of the display data based on the analysis of the monitored light, compressing the adjusted display data, and sending (S84) it for display on the display screen. In one embodiment, the one or more predetermined parameters includes one or more colour thresholds and monitoring the light comprises monitoring relative levels of different colours in the monitored light.

Abstract: A decompression system includes a first memory including a first write port configured to receive decompressed data from a decompressor, and a first read port configured to receive a back-reference read request, the first memory being configured to output the decompressed data to the decompressor in response to receiving the back-reference read request at the first read port, and a second memory including a second write port electrically coupled to the first write port and configured to receive the decompressed data, the second memory being configured to buffer the decompressed data for retrieval by a receiver.

Abstract: Gaze tracking data is analyzed to determine one or more regions of interest within an image of a video stream. The video stream data is selectively scaled so that sections within the regions of interest maintain high resolution while areas not within the region of interest are down-scaled to reduce bandwidth cost of transmission. A scheme for reduction of motion sickness by reducing the size of the high resolution area is also claimed.

Abstract: Aspects of the present disclosure involve systems and methods for performing operations comprising receiving, with a messaging application, user input to access a graphical image modification feature of the messaging application; in response to receiving, causing display of a video; accessing a first configuration rule of a plurality of configuration rules that associates a first device property rule with the graphical image modification feature of the messaging application; determining that the first configuration rule is satisfied by a first property of the client device; and in response to determining that the first configuration rule is satisfied by the first property of the client device, causing display of a first plurality of graphical image modification options each associated with performing a different modification to the video.

Abstract: A storage system that includes an in-line hardware accelerator, a solid-state drive (SSD) unit, a central processing unit (CPU), a volatile memory module, and an accelerator memory module that is coupled to the in-line hardware accelerator or belongs to the in-line hardware accelerator; wherein the in-line hardware accelerator is directly coupled to the SSD unit, the volatile memory and the non-volatile memory; wherein the CPU is directly coupled to the volatile memory and to the non-volatile memory; wherein the in-line hardware accelerator is configured to manage access to the SSD unit; wherein the in-line accelerator is configured to retrieve data stored in the volatile memory module and the non-volatile memory module without involving the CPU.

Abstract: Example methods, apparatus, systems and articles of manufacture (e.g., non-transitory physical storage media) to implement foveated image rendering for head-mounted displays device are disclosed herein. Example head-mounted display devices disclosed herein include a frame buffer to store first and second image data for an image frame, the first image data having a first resolution and the second image data having a second resolution lower than the first resolution, the first image data and the second image data obtained from a host device via a data interface. Disclosed example head-mounted display devices also include a device controller to up-sample the second image data based on first metadata from the host device to generate up-sampled second image data having the first resolution, and combine the first image data and the up-sampled second image data based on second metadata from the host device to render a foveated image frame on a display.

Abstract: Embodiments provide for a graphics processing apparatus including a cache memory and logic coupled to the cache memory to compress color data output from the first cache memory. In one embodiment the cache memory is a render cache. In one embodiment the cache memory is a victim data cache. In one embodiment the first cache memory is a render cache coupled to a victim data cache and logic is configured to compress color data evicted from the render cache and the victim data cache. The compression can include a target compression ratio to which the data is to be compressed.

Abstract: Systems and methods are configured to adjust the timing of source frame compression in response to fluctuations in a variable frame rate at which source frames are rendered.

Abstract: A method of stress compensation in a display includes converting a stress profile for a slice of the display from a first format to a second format based on a conversion ratio; transforming the converted stress profile for the slice of the display, with a first transformation, to form a compressed transformed stress profile; decompressing the compressed transformed stress profile to form a decompressed transformed stress profile; and transforming the decompressed transformed stress profile, with a second transformation, to form a decompressed stress profile, the second transformation being an inverse of the first transformation.

Abstract: Techniques of this disclosure may include ways to control the amount of graphics data a graphics processing unit (GPU) renders. The GPU may render graphics data for image content that changed from frame-to-frame rather than graphics data for image content that changed and did not change. To display the image content, processing circuitry may map locations of where the graphics data is stored to lines in the image content allowing for the GPU to store the graphics data in arbitrary locations of an application buffer.

Abstract: A method controls execution of an intended application. One or more processors receive a first formatted character string that is in a first format. The processor(s) determine which application from multiple applications is an intended application that is to use the first formatted character string as a basis of an input to the intended application. The processor(s) convert the first formatted character string into a second formatted character string by applying a second format that is used by the intended application, and then execute the intended application by using the second formatted character string as an input to the intended application.

Abstract: A method of identifying a show in a video that has been filmed by a spectator's camera includes selecting a color set to identify the show, recording, in a database, a show identifier associated with a code for the color set, before the show is filmed, inserting, within a scene in which the show is performed, a display-panel set, causing the display-panel set to display the color set without considering scene lighting, thereby raising a likelihood that the camera will capture the displayed color set, once the camera has begun filming a video of the scene, causing a tracer to download an image of the video, to search within the image for the color set associated by the database with the show identifier, and to respond to finding the color set by identifying the filmed show with the aid of the show identifier.

Abstract: Semiconductor devices with controllers under stacks of semiconductor packages and associated systems and methods are disclosed herein. In one embodiment, a semiconductor device includes a package substrate, a controller attached to the package substrate, and at least two semiconductor packages disposed over the controller. Each semiconductor package includes a plurality of semiconductor dies. The semiconductor device further includes an encapsulant material encapsulating the controller and the at least two semiconductor packages.

Abstract: Innovations in content mastering operations performed during playback of high dynamic range (“HDR”) video on a display device are described. When content mastering is performed during playback on a display device, a video playback system can use details retained for input HDR video (e.g., retained in metadata) and the properties of the display device to improve the perceptual quality of the HDR video as shown on that display device. For example, the video playback system can use an energy-preserving bloom operator to make bright highlights “bloom” into adjacent areas, thereby accentuating the bright highlights in the HDR video while operating within the constraints of the display device. The video playback system can also perform various other types of operations when content mastering is deferred until playback, including application of a lens flare operator as well as alternative tone mapping operators and alternative color gamut mapping operators selected according to metadata.

Abstract: Systems, methods and apparatuses may provide for technology to reduce rendering overhead associated with light field displays. The technology may conduct data formatting, re-projection, foveation, tile binning and/or image warping operations with respect to a plurality of display planes in a light field display.

Abstract: A method of video coding using palette coding mode for color video in a non-444 color format is disclosed. The non-444 video data can be converted to the 444 format so that the palette designed for the 444 format can be used for the non-444 video data. For the 4:2:0 color format, each chroma sample collocated with corresponding quad luma samples is padded with three chroma samples to form converted quad chroma samples. For the 4:2:0 color format, if a current pixel does not belong to the color palette, one or more escape values are signaled for the current pixel and the escape value associated with the chroma sample of the current pixel for each color component is not signaled for three of the quad chroma sample locations collocated with corresponding quad luma samples.

Abstract: Embodiments provide for a graphics processing apparatus including a cache memory and logic coupled to the cache memory to compress color data output from the first cache memory. In one embodiment the cache memory is a render cache. In one embodiment the cache memory is a victim data cache. In one embodiment the first cache memory is a render cache coupled to a victim data cache and logic is configured to compress color data evicted from the render cache and the victim data cache. The compression can include a target compression ratio to which the data is to be compressed.

Abstract: A method and apparatus are provided for compressing depth buffer data in a three dimensional computer graphics system. The depth buffer data is divided into a plurality of rectangular tiles corresponding to rectangular areas in an associated image. The number of starting point locations in a tile are identified and a difference in depth value determined between each starting point and depth values of each of at least two further locations. Using this information depth values are predicted at a plurality of other locations in the tile and where these predicated values substantially match an actual depth value at location is assigned to a plane associated with respective starting point. Starting point location depth value difference data and plane assignment data for each tile and locations in the tile not assigned to a plane, then stored.

Abstract: According to the invention, a system for presenting graphics on a display device is disclosed. The system may include an eye tracking device for determining a gaze point of a user on a display device. The system may also include a graphics processing device for causing graphics to be displayed on the display device. The graphics displayed on the display device may be modified such that the graphics in an area including the gaze point of the user have at least one modified parameter relative to graphics outside the area. The size of the area may be based at least in part on an amount of noise in the gaze point over time and at least one other secondary factor.

Abstract: The use of the three-dimensional DCT as a key compression technology requires development of an entirely new quantizing mechanism. The embodiment described herein uses a Human Visual Model to develop quantizers based on a combination of descriptive characteristics of the video source, enabling independent derivation of said quantizers in both encoder and decoder sides of the compression and playback process.

Abstract: A system and method for decompressing compressed data (e.g., in a frame buffer) and optionally recompressing the data. The method includes determining a portion of an image to be accessed from a memory and sending a conditional read corresponding to the portion of the image. In response to the conditional read, an indicator operable to indicate that the portion of the image is uncompressed may be received. If the portion of the image is compressed, in response to the conditional read, compressed data corresponding to the portion of the image is received. In response to receiving the compressed data, the compressed data is uncompressed into uncompressed data. The uncompressed data may then be written to the memory corresponding to the portion of the image. The uncompressed data may then be in-place compressed for or during subsequent processing.

Abstract: Apparatuses, methods, systems, and program products are disclosed for managing storage of digital content. An eligibility module determines one or more content elements that are eligible for compression. A content element is determined to be eligible for compression based on one or more characteristics of the content element. A rate module determines a compression rate for each of the one or more content elements. The compression rate comprises an amount of compression to be applied to a content element. The amount of compression to be applied to the content element is determined as a function of one or more characteristics of the content element. A compression module compresses each of the one or more eligible content elements according to the determined compression rate.

Abstract: A graphics processing pipeline determines whether respective graphics processing operations, such as respective blends, respective depth tests, etc., to be performed at a stage of the graphics processing pipeline would produce the same result for each sampling point of a set of plural sampling points represented by a fragment being processed by the graphics processing pipeline. If it is determined that respective graphics processing operations would produce the same result for each of the sampling points, then only a single instance of the graphics processing operation is performed and the result of that graphics processing operation is associated with each of the sampling points. The number of instances of the graphics processing operations needed to process the set of plural sampling points which the fragment represents is reduced in comparison to conventional multisampling graphics processing techniques which perform graphics processing operations for fragments on a “per sample” basis.

Abstract: Data destined for memory, i.e., data that was evicted at some level in the cache hierarchy is intercepted and subjected to compression before being sent to memory. Thereby, when the compression is successful, the memory bandwidth requirement is reduced, potentially resulting in higher performance and/or energy efficiency in some embodiments.

Abstract: A data driver including a power control part configured to control power according to mode signal determined based on an input image, a digital to analog converting part configured to convert a digital data signal into an analog data voltage, a buffering part configured to buffer the data voltage, a first switching part configured to apply the data voltage to a data line in a normal mode, when turned on, and a second switching part configured to apply a blank voltage to the data line in a blank period of a low frequency mode, when turned on.

Abstract: A method for encoding high dynamic range (HDR) images involves providing a lower dynamic range (LDR) image, generating a prediction function for estimating the values for pixels in the HDR image based on the values of corresponding pixels in the LDR image, and obtaining a residual frame based on differences between the pixel values of the HDR image and estimated pixel values. The LDR image, prediction function and residual frame can all be encoded in data from which either the LDR image of HDR image can be recreated.

Abstract: Apparatus and associated methods relate to compressing a sequence of binary data by encoding difference values between adjacent data in the sequence. For each datum in the sequence, the difference value may be determined by comparing the datum with an immediately preceding datum. For data, the determined difference may be expressible using few bits. The determined difference is encoded in subwords having subword lengths selected from a set of predetermined subword lengths. The selected subword lengths may be a minimum one of the set of predetermined subword sizes that is capable of representing the difference between the adjacent data. A size tag is generated for each subword encoded. The size tag identifies the size of the subword selected and used for encoding the determined difference. The difference may be encoded as a mathematical difference or as a bit pattern difference.

Abstract: Embodiments described herein provide systems and methods for managing quality level of video transmission and storage. In a particular embodiment, a method provides receiving a video at a received quality level in a video storage system and determining a characteristic of the video. The method further provides determining whether the characteristic satisfies a criteria and, if the characteristic satisfies the criteria, storing the video at a quality level associated with the criteria.

Abstract: It may be determined that a payment reader requires a firmware update, which may be transmitted to the payment reader as compressed firmware update blocks. The payment reader may receive a first portion of set of the compressed firmware update blocks. The payment reader may decompress the first portion and determine a partial firmware offset associated with the first portion. If the firmware update is incomplete, the payment reader may transmit that partial firmware offset to a second device. Based on this partial of firmware offset and an offset table that associates compressed firmware offsets with decompressed firmware offsets, the payment reader receive a second portion of compressed firmware update blocks to send to the payment reader. The payment reader may determine that entire update has been received and update its firmware.

Abstract: Embodiments of the present invention allow a data segment stored in a compressed format to be directly staged into a cache allowing access to the data segment. In operation, a request to access a data segment is received. In response, whether the data segment is stored in a compressed format is determined. If the data segment is stored in the compressed format, the data segment is staged directly into a cache.

Abstract: Systems and methods provide for non-multisampled anti-aliasing for clipping paths, in which a non-multisampled texture is employed to store anti-aliasing data. In some configurations, clipping paths applied to an input object are processed successively using a non-multi-sampled buffer and non-multisampled texture. Each clipping path is processed by incrementing a stencil buffer value for each pixel covered by the clipping path, computing clipping path coverage data, and storing the clipping path coverage data in the non-multisampled texture. An object is rendered by performing a stencil test and multiplying color values for retained pixels by corresponding texture values from the non-multisampled texture to provide anti-aliasing. Further configurations operate without a stencil buffer but employ a logical stack of non-multisampled textures, one for each clipping path.