Intel Corporation

Karthik Vaidyanathan

Prasoonkumar Surti

Abhishek R Appu

Joydeep Ray

Murali Ramadoss

Michael Apodaca

Carsten Benthin

Abhishek Venkatesh

Travis T Schluessler

Gabor Liktor

John H Feit

Altug Koker

Adam T Lake

Jonathan Kennedy

Atsuo Kuwahara

Sven Woop

Balaji Vembu

Subramaniam M Maiyuran

Jeffery S Boles

Devan Burke

Hugues Labbe

Sameer Kp

Marco Salvi

Philip R Laws

Michael Doyle

Narayan Srinivasa

Kamal Sinha

Justin E Gottschlich

Feng Chen

Eriko Nurvitadhi

Linda L Hurd

Sara S Baghsorkhi

Ben J Ashbaugh

Nicolas C Galoppo Von Borries

Rajkishore Barik

Dukhwan Kim

Nadathur Rajagopalan Satish

Tsung-Han Lin

Wenyin Fu

Josh B Mastronarde

Farshad Akhbari

Scott R Janus

Karol A Szerszen

Nikos Kaburlasos

Jacek Kwiatkowski

Slawomir Grajewski

Vasanth Ranganathan

Thomas F Raoux

Elmoustapha Ould-Ahmed-Vall

Kai Xiao

Robert M Toth

Tomer Bar-On

Andrew T Lauritzen

David M Cimini

Jefferson Amstutz

Joshua D Barczak

Eric G Liskay

Tomas G Akenine-Moller

Carl Jacob Munkberg

John G Gierach

Franz Petrik Clarberg

Tomasz Janczak

Eric J Hoekstra

Carson Brownlee

Alexey Supikov

Saikat Mandal

Joshua Barczak

Ankur N Shah

Jon N Hasselgren

Peter L Doyle

David Robert Baldwin

Louis Feng

Shashank Lakshminarayana

Sudhakar Kamma

Sreenivas Kothandaraman

Abhishek Rhisheekesan

Won-Jong Lee

Rajasekhar Pantangi

Priyanka Ladda

Pawel Majewski

Holger Gruen

Stephen Junkins

Sungye Kim

Vikranth Vemulapalli

Aaron Eliot Lefohn

Rahul P Sathe

Michael J Norris

Pradeep Ramani

Tim Foley

Valentin Fuetterling

Sameer K P

Subramanian Maiyuran

Manu Mathew Thomas

Bret G Alfieri

One embodiment provides a graphics processor comprising a set of processing resources configured to perform a supersampling operation via a mixed precision convolutional neural network, the set of processing resources including circuitry configured to receive, at an input block of a neural network model, history data, velocity data, and current frame data, pre-process the history data, velocity data, and current frame data to generate pre-processed data, provide the pre-processed data to a feature extraction network of the neural network model, process the pre-processed data at the feature extraction network via one or more encoder stages and one or more decoder stages, and generate an output image via an output block of the neural network model via direct reconstruction or kernel prediction.

Temporally amortized supersampling using a mixed precision convolutional neural network

Apparatus and method for lossy displaced mesh compression. For example, one embodiment of an apparatus comprises: displacement mapping circuitry/logic to generate an original displacement-mapped mesh by performing a displacement mapping of a plurality of vertices of a base subdivision mesh; and mesh compression circuitry/logic to compress the original displacement-mapped mesh, the mesh compression circuitry/logic comprising a quantizer to quantize the displacement mapping of the plurality of vertices in view of a base mesh to generate a displacement array.

Apparatus and method for displaced mesh compression

Systems, apparatuses and methods may provide for technology that determines a stencil value and uses the stencil value to control, via a stencil buffer, a coarse pixel size of a graphics pipeline. Additionally, the stencil value may include a first range of bits defining a first dimension of the coarse pixel size and a second range of bits defining a second dimension of the coarse pixel size. In one example, the coarse pixel size is controlled for a plurality of pixels on a per pixel basis.

Controlling coarse pixel size from a stencil buffer

An apparatus and method to execute ray tracing instructions. For example, one embodiment of an apparatus comprises execution circuitry to execute a dequantize instruction to convert a plurality of quantized data values to a plurality of dequantized data values, the dequantize instruction including a first source operand to identify a plurality of packed quantized data values in a source register and a destination operand to identify a destination register in which to store a plurality of packed dequantized data values, wherein the execution circuitry is to convert each packed quantized data value in the source register to a floating point value, to multiply the floating point value by a first value to generate a first product and to add the first product to a second value to generate a dequantized data value, and to store the dequantized data value in a packed data element location in the destination register.

Apparatus and method for ray tracing instruction processing and execution

An embodiment of an electronic processing system may include an application processor, persistent storage media communicatively coupled to the application processor, a graphics subsystem communicatively coupled to the application processor, a sense engine communicatively coupled to the graphics subsystem to provide sensed information, a focus engine communicatively coupled to the sense engine and the graphics subsystem to provide focus information, a motion engine communicatively coupled to the sense engine, the focus engine, and the graphics subsystem to provide motion information, and a motion biased foveated renderer communicatively coupled to the motion engine, the focus engine, the sense engine to adjust one or more parameters of the graphics subsystem based on one or more of the sense information, the focus information, and the motion information. Other embodiments are disclosed and claimed.

Motion biased foveated renderer

Apparatus and method for grouping rays based on quantized ray directions. For example, one embodiment of an apparatus comprises: An apparatus comprising: a ray generator to generate a plurality of rays; ray direction evaluation circuitry/logic to generate approximate ray direction data for each of the plurality of rays; ray sorting circuitry/logic to sort the rays into a plurality of ray queues based, at least in part, on the approximate ray direction data.

Apparatus and method for quantized convergent direction-based ray sorting

Embodiments provide mechanisms to facilitate compute operations for deep neural networks. One embodiment comprises a graphics processing unit comprising one or more multiprocessors, at least one of the one or more multiprocessors including a register file to store a plurality of different types of operands and a plurality of processing cores. The plurality of processing cores includes a first set of processing cores of a first type and a second set of processing cores of a second type. The first set of processing cores are associated with a first memory channel and the second set of processing cores are associated with a second memory channel.

Compute optimization mechanism for deep neural networks

Apparatus and method for speculative execution of hit and intersection shaders on programmable ray tracing architectures. For example, one embodiment of an apparatus comprises: single-instruction multiple-data (SIMD) or single-instruction multiple-thread (SIMT) execution units (EUs) to execute shaders; and ray tracing circuitry to execute a ray traversal thread, the ray tracing engine comprising: traversal/intersection circuitry, responsive to the traversal thread, to traverse a ray through an acceleration data structure comprising a plurality of hierarchically arranged nodes and to intersect the ray with a primitive contained within at least one of the nodes; and shader deferral circuitry to defer and aggregate multiple shader invocations resulting from the traversal thread until a particular triggering event is detected, wherein the multiple shaders are to be dispatched on the EUs in a single shader batch upon detection of the triggering event.

Speculative execution of hit and intersection shaders on programmable ray tracing architectures

Embodiments are generally directed to compression for compression for sparse data structures utilizing mode search approximation. An embodiment of an apparatus includes one or more processors including a graphics processor to process data; and a memory for storage of data, including compressed data. The one or more processors are to provide for compression of a data structure, including identification of a mode in the data structure, the data structure including a plurality of values and the mode being a most repeated value in a data structure, wherein identification of the mode includes application of a mode approximation operation, and encoding of an output vector to include the identified mode, a significance map to indicate locations at which the mode is present in the data structure, and remaining uncompressed data from the data structure.

Compression for sparse data structures utilizing mode search approximation

Apparatus and method for stack access throttling for synchronous ray tracing. For example, one embodiment of an apparatus comprises: ray tracing acceleration hardware to manage active ray tracing stack allocations to ensure that a size of the active ray tracing stack allocations remains within a threshold; and an execution unit to execute a thread to explicitly request a new ray tracing stack allocation from the ray tracing acceleration hardware, the ray tracing acceleration hardware to permit the new ray tracing stack allocation if the size of the active ray tracing stack allocations will remain within the threshold after permitting the new ray tracing stack allocation.

Stack access throttling for synchronous ray tracing

Apparatus and method for efficient BVH construction. For example, one embodiment of an apparatus comprises: a memory to store graphics data for a scene including a plurality of primitives in a scene at a first precision; a geometry quantizer to read vertices of the primitives at the first precision and to adaptively quantize the vertices of the primitives to a second precision associated with a first local coordinate grid of a first BVH node positioned within a global coordinate grid, the second precision lower than the first precision; a BVH builder to determine coordinates of child nodes of the first BVH node by performing non-spatial-split binning or spatial-split binning for the first BVH node using primitives associated with the first BVH node, the BVH builder to determine final coordinates for the child nodes based, at least in part, on an evaluation of surface areas of different bounding boxes generated for each of the child node.

Apparatus and method for reduced precision bounding volume hierarchy construction

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.

Apparatus and method for efficient graphics processing including ray tracing

Apparatus and method for more precise level-of-details transitions. For example one embodiment includes a graphics processor comprising: ray traversal hardware logic to traverse a ray through an acceleration structure to determine intersections between the ray and one or more object instances; and a level of detail selector to: set an instance comparison mask associated with an object instance to a first level of detail (LOD), the instance comparison mask comprising an N-bit value and one or more bits to indicate a type of comparison operation, compare a value from a ray mask with the N-bit value in accordance with the type of comparison operation to generate a comparison result, and determine whether to use the first LOD or a second LOD to render one or more pixels in accordance with the comparison result.

Apparatus and method for high quality ray-traced level of detail transitions

Cluster of acceleration engines to accelerate intersections. For example, one embodiment of an apparatus comprises: a set of graphics cores to execute a first set of instructions of a primary graphics thread; a scalar cluster comprising a plurality of scalar execution engines; and a communication fabric interconnecting the set of graphics cores and the scalar cluster; the set of graphics cores to offload execution of a second set of instructions associated with ray traversal and/or intersection operations to the scalar cluster; the scalar cluster comprising a plurality of local memories, each local memory associated with one of the scalar execution engines, wherein each local memory is to store a portion of a hierarchical acceleration data structure required by an associated scalar execution engine to execute one or more of the second set of instructions; the plurality of scalar execution engines to store results of the execution of the second set of instructions in a memory accessible by the set of graphics cores; wherein the set of graphics cores are to process the results within the primary graphics thread.

Cluster of scalar engines to accelerate intersection in leaf node

An apparatus to facilitate compression of memory data is disclosed. The apparatus comprises one or more processors to receive uncompressed data, adapt a format of the uncompressed data to a compression format, perform a color transformation from a first color space to a second color space, perform a residual computation to generate residual data, compress the residual data via entropy encoding to generate compressed data and packing the compressed data.

Unified memory compression mechanism

Growing community of inventors

Berkeley, CA, United States of America

Karthik Vaidyanathan