Microsoft Technology Licensing, LLC

Anuj Bharat Gosalia

Steve Pronovost

David Floyd Aronson

Charles Neil Boyd

David John Martin

Craig C Peeper

Daniel Kurt Baker

Loren McQuade

Michael A Toelle

Jeff M J Noyle

Bryan L Langley

Hang Li

Leonardo Esteban Blanco

Amar Dinesh Patel

Stephen Estrop

Silvana Patricia Moncayo

David R Blythe

Max Alan McMullen

Gregory David Swedberg

Benjamin C Constable

Joseph Cox Ballantyne

Daniel Wood

Joshua Warren Priestley

Ameet Arun Chitre

Iouri Tarassov

Jay Senior

Aleksandar Antonijevic

Kan Qiu

Hideyuki Nagase

Imran Majid

Mary Luo

John (Mingtzong) Lee

Anantha R Kancheria

One or more techniques and/or systems are provided for operating a graphics processing unit (GPU). A sensor of a computing device may collect sensor input data (e.g., camera input, touch input, video input, etc.), which may be provided to the GPU. An input process within the GPU may be invoked to process the sensor input data to generate a result that may be retained within GPU accessible memory (e.g., a touch sensor process may generate a gesture result based upon touch input from a touch panel of the computing device). An output process within the GPU may be invoked to utilize the result within the GPU accessible memory, for display rendering. In this way, latency between user input and display rendering may be mitigated by streamlining processing on the GPU by mitigating transmission of data between the GPU and a CPU of the computing device for display rendering.

GPU operation

Atlasing and virtual surface techniques are described. In one or more implementations, virtual surface functionality is exposed by an operating system for access by one or more applications of the computing device. A virtual surface is created in response to a request from the one or more applications to be used to render visuals for display by a display device. The virtual surface is allocated in memory of the computing device by the exposed virtual surface functionality to have an area that is larger than an area to be used to display the visuals from the one or more applications.

Atlasing and virtual surfaces

Techniques for minimizing coprocessor 'starvation,' and for effectively scheduling processing in a coprocessor for greater efficiency and power. A run list is provided allowing a coprocessor to switch from one task to the next, without waiting for CPU intervention. A method called 'surface faulting' allows a coprocessor to fault at the beginning of a large task rather than somewhere in the middle of the task. DMA control instructions, namely a &#x201c;fence,&#x201d; a &#x201c;trap&#x201d; and a &#x201c;enable/disable context switching,&#x201d; can be inserted into a processing stream to cause a coprocessor to perform tasks that enhance coprocessor efficiency and power. These instructions can also be used to build high-level synchronization objects. Finally, a &#x201c;flip&#x201d; technique is described that can switch a base reference for a display from one location to another, thereby changing the entire display surface.

Building a run list for a coprocessor based on rules when the coprocessor switches from one context to another context

A high level shader language compiler incorporates transforms to optimize shader code for graphics processing hardware. An instruction reordering transform determines instruction encapsulations of dependent instructions that reduce concurrent register usage by the shader. A phase pulling transform re-organizes the shader's instructions into phases that reduce a measure of depth of texture loads. A register assigning transform assigns registers to lower register usage by the shader.

Optimizing compiler transforms for a high level shader language

Systems and methods are provided for scheduling the processing of a coprocessor whereby applications can submit tasks to a scheduler, and the scheduler can determine how much processing each application is entitled to as well as an order for processing. In connection with this process, tasks that require processing can be stored in physical memory or in virtual memory that is managed by a memory manager. The invention also provides various techniques of determining whether a particular task is ready for processing. A 'run list' may be employed to ensure that the coprocessor does not waste time between tasks or after an interruption. The invention also provides techniques for ensuring the security of a computer system, by not allowing applications to modify portions of memory that are integral to maintaining the proper functioning of system operations.

Multithreaded kernel for graphics processing unit

Systems and methods that independently control divided and/or isolated processing resources of a Graphical Processing Unit (GPU). Synchronization primitives for processing are shared among such resources to process interaction with the engines and their associated different requirements (e.g. different language). Accordingly, independent threads can be created against particular nodes (e.g., a video engine node, 3D engine node), wherein multiple engines can exist under a single node, and independent control can subsequently be exerted upon the plurality of engines associated with the GPU.

Parallel engine support in display driver model

Video frame buffers are controlled using a sequence of new-frame-indicators (e.g., FLIP) and no-new-frame-indicators (e.g., NOFLIP) in a frame indicator queue that is accessed with each display refresh. Video samples are loaded into a chain of video frame buffers that is 'rotated' during the vertical blanking signal of the display to swap an old frame buffer out for a new frame buffer. The rotations of the frame buffer chain are controlled based on the frame indicators in the frame indicator queue to present new video samples to the display in a regular pattern, thereby providing smooth video playback.

Frame buffer control for smooth video display

A video memory manager manages and virtualizes memory so that an application or multiple applications can utilize both system memory and local video memory in processing graphics. The video memory manager allocates memory in either the system memory or the local video memory as appropriate. The video memory manager may also manage the system memory accessible to the graphics processing unit via an aperture of the graphics processing unit. The video memory manager may evict memory from the local video memory as appropriate, thereby freeing a portion of local video memory use by other applications. In this manner, a graphics processing unit and its local video memory may be more readily shared by multiple applications.

Video memory management

Systems and methods for scheduling coprocessing resources in a computing system are provided without redesigning the coprocessor. In various embodiments, a system of preemptive multitasking is provided achieving benefits over cooperative multitasking by any one or more of (1) executing rendering commands sent to the coprocessor in a different order than they were submitted by applications; (2) preempting the coprocessor during scheduling of non-interruptible hardware; (3) allowing user mode drivers to build work items using command buffers in a way that does not compromise security; (4) preparing DMA buffers for execution while the coprocessor is busy executing a previously prepared DMA buffer; (5) resuming interrupted DMA buffers; and (6) reducing the amount of memory needed to run translated DMA buffers.

Systems and methods for scheduling coprocessor resources in a computing system

Systems and methods for enhancing performance of a coprocessor

Complex computer graphics forms and motions can be constructed either by hand or with motion or geometry capture technologies, once they are created, they are difficult to modify, particularly at runtime. Interpolation provides a way to leverage artist-generated source material. Methodologies for efficient runtime interpolation between multiple forms or multiple motion segments enables computers to perform more realistic animation in real-time. Shape interpolation is applied to predefined figures to create smoothly skinned figures that deform in natural ways. Predefined figures are selected using a search technique that reduces the amount of interpolation required to produce real-time animation.

Methods and system for general skinning via hardware accelerators

Growing community of inventors

Redmond, WA, United States of America

Anuj Bharat Gosalia