SYSTEM AND METHOD FOR EFFICIENT MULTI-GPU EXECUTION OF KERNELS BY REGION BASED DEPENDENCIES

    公开(公告)号:WO2021222377A1

    公开(公告)日:2021-11-04

    申请号:PCT/US2021/029592

    申请日:2021-04-28

    Abstract: Methods for graphics processing are provided. One example method includes executing a plurality of kernels using a plurality of graphics processing units (GPUs), wherein responsibility for executing a corresponding kernel is divided into one or more portions each of which being assigned to a corresponding GPU. The method includes generating a plurality of dependency data at a first kernel as each of a first plurality of portions of the first kernel completes processing. The method includes checking dependency data from one or more portions of the first kernel prior to execution of a portion of a second kernel. The method includes delaying execution of the portion of the second kernel as long as the corresponding dependency data of the first kernel has not been met.

    SYSTEM AND METHOD FOR MULTI-TENANT IMPLEMENTATION OF GRAPHICS PROCESSING UNIT

    公开(公告)号:WO2021066974A1

    公开(公告)日:2021-04-08

    申请号:PCT/US2020/048641

    申请日:2020-08-28

    Inventor: CERNY, Mark E.

    Abstract: A method for graphics processing, wherein a graphics processing unit (GPU) resource is allocated among applications, such that each application is allocated a set of time slices. Commands of draw calls are loaded to rendering command buffers in order to render an image frame for a first application. The commands are processed by the GPU resource within a first time slice allocated to the first application. The method including determining at least one command has not been executed at an end of the first time slice. The method including halting execution of commands, wherein remaining one or more commands are not processed in the first time slice. A GPU configuration is preserved for the commands after processing a last executed command, the GPU configuration used when processing in a second time slice the remaining commands.

    MICROPHONE ON CONTROLLER WITH TOUCHPAD TO TAKE IN AUDIO SWIPE FEATURE DATA

    公开(公告)号:WO2021021414A1

    公开(公告)日:2021-02-04

    申请号:PCT/US2020/041422

    申请日:2020-07-09

    Abstract: A game controller (202) includes a touchpad (204) that a user, viewing a virtual keyboard (216) on a screen, can soft-touch (300) to move a cursor on the screen and then hard-touch (302) to move the cursor and also send location data to a processor for inputting a letter from the virtual keyboard. A microphone (218) on the touchpad can be used to receive voice signals for training a machine learning module to predict a next letter or next word, or to insert special characters/punctuations/ graphics such as "smileys" during the swipe, or to indicate a tone of a Chinese character while typing with Chinese Pinyin.

    SELF-HEALING MACHINE LEARNING SYSTEM FOR TRANSFORMED DATA

    公开(公告)号:WO2021011154A1

    公开(公告)日:2021-01-21

    申请号:PCT/US2020/038965

    申请日:2020-06-22

    Abstract: An end-to-end cloud-based machine learning platform providing personalized game player experiences. Data lineage (102) is generated for all transformed data for generating feature ETLs, and for training machine learning models (304). That data is used to understand the performance of off-line and online recommender systems for the personalization of the game player experiences. To that end, the platform pipeline provides the life cycle of the transformed data to a self-healing system that compare it to the life cycle of the user interactions. By comparing the two life cycles, the self-healing system can automatically provide a diagnostic (e.g., 510), and it can also automatically provide an action (e.g., 512) if the performance of the model predictions has changed over time.

    ASSET AWARE COMPUTING ARCHITECTURE FOR GRAPHICS PROCESSING

    公开(公告)号:WO2021002949A1

    公开(公告)日:2021-01-07

    申请号:PCT/US2020/032954

    申请日:2020-05-14

    Inventor: CERNY, Mark, E.

    Abstract: A method for executing a game by a computing system that uses a central processing unit (CPU) and graphics processing unit (GPU) for generating video frames. A draw call is generated for a video frame by the CPU. At bind time, i.e. writing of the GPU commands by the CPU using a GPU API, asset aware data (AAD) is written to the command buffer, and loading of one or more level of detail (LOD) data from an asset store to system memory is requested. The GPU executes the draw call for the frame using LOD data written to the system memory, the GPU using at least a minimum of LOD data based on the AAD. Additionally, the GPU uses information regarding the LOD load state when executing the draw call, in order to avoid access to LODs not yet loaded.

    ERROR DE-EMPHASIS IN LIVE STREAMING
    90.
    发明申请

    公开(公告)号:WO2020172026A1

    公开(公告)日:2020-08-27

    申请号:PCT/US2020/017978

    申请日:2020-02-12

    Abstract: A Reductive Edging device and computer program product. The Reductive Edging device has a processor, memory coupled to the processor and non-transitory instructions embedded in the memory that when executed by the processor causes the device to perform a method for reductive edging. The method includes, prior to sending a media segment to a client device, performing error de-emphasis operations and sending information associated with error de-emphasis.

Patent Agency Ranking