Abstract: Described herein is a technique for performing ray-triangle intersection without a floating point division unit. A division unit would be useful for a straightforward implementation of a certain type of ray-triangle intersection test that is useful in ray tracing operations. This certain type of ray-triangle intersection test includes a step that transforms the coordinate system into the viewspace of the ray, thereby reducing the problem of intersection to one of 2D triangle rasterization. However, a straightforward implementation of this transformation requires floating point division, as the transformation utilizes a shear operation to set the coordinate system such that the magnitudes of the ray direction on two of the axes are zero. Instead of using the most straightforward implementation of this transform, the technique described herein scales the entire coordinate system by the magnitude of the ray direction in the axis that is the denominator of the shear ratio, removing division.

Abstract: Bus protocol features are provided for chaining memory access requests on a high speed interconnect bus, allowing for reduced signaling overhead. Multiple memory request messages are received over a bus. A first message has a source identifier, a target identifier, a first address, and first payload data. The first payload data is stored in a memory at locations indicated by the first address. Within a selected second one of the request messages, a chaining indicator is received associated with the first request message and second payload data. The second request message does not include an address. Based on the chaining indicator, a second address for which memory access is requested is calculated based on the first address. The second payload data is stored in the memory at locations indicated by the second address.

Abstract: Described herein is a technique for performing ray tracing. According to this technique, instead of executing intersection and/or any hit shaders during traversal of an acceleration structure to determine the closest hit for a ray, an acceleration structure is fully traversed in an invocation of a shader program, and the closest intersection with a triangle is recorded in a data structure associated with the material of the triangle. Later, a scheduler launches waves by grouping together multiple data items associated with the same material. The rays processed by that wave are processed with a continuation ray, rather than the full original ray. A continuation ray starts from the previous point of intersection and extends in the direction of the original ray. These steps help counter divergence that would occur if a single shader program that inlined the intersection and any hit shaders were executed.

Abstract: A link controller, method, and data processing platform are provided with dual-protocol capability. The link controller includes a physical layer circuit for providing a data lane over a communication link, a first data link layer controller which operates according to a first protocol, and a second data link layer controller which operates according to a second protocol. A multiplexer/demultiplexer selectively connects both data link layer controllers to the physical layer circuit. A link training and status state machine (LTSSM) selectively controls the physical layer circuit to transmit and receive first training ordered sets over the data lane, and inside the training ordered sets, transmit and receive alternative protocol negotiation information over the data lane. In response to receiving the alternative protocol negotiation information, the LTSSM causes the multiplexer/demultiplexer to selectively connect the physical layer circuit to the second data link layer controller.

Abstract: A processing device is provided which includes memory and at least one processor. The memory includes main memory and cache memory in communication with the main memory via a link. The at least one processor is configured to receive a request for a cache line and read the cache line from main memory. The at least one processor is also configured to compress the cache line according to a compression algorithm and, when the compressed cache line includes at least one byte predicted not to be accessed, drop the at least one byte from the compressed cache line based on whether the compression algorithm is determined to successfully compress the cache line according to a compression parameter.

Abstract: Described herein is a technique for performing ray-triangle intersection test in a manner that produces watertight results. The technique involves translating the coordinates of the triangle such that the origin is at the origin of the ray. The technique involves projecting the coordinate system into the viewspace of the ray. The technique then involves calculating barycentric coordinates and interpolating the barycentric coordinates to get a time of intersect. The signs of the barycentric coordinates indicate whether a hit occurs. The above calculations are performed with a non-directed floating point rounding mode to provide watertightness. A non-directed rounding mode is one in which the mantissa of a rounded number is rounded in a manner that is not dependent on the sign of the number.

Abstract: A link controller includes a Peripheral Component Interconnect Express (PCIe) physical layer circuit for coupling to a communication link and providing a data path over the communication link, a first data link layer controller which operates according to a PCIe protocol, and a second data link layer controller which operates according to a Gen-Z protocol. A multiplexer-demultiplexer selectively connects both data link layer controllers to the PCIe physical layer circuit. A protocol translation circuit is coupled between the multiplexer-demultiplexer and the second data link layer controller, the protocol translation circuit receiving traffic data from the second data link layer controller in a Gen-Z format, encapsulating the Gen-Z format in a PCIe format, and passing traffic data to the multiplexer-demultiplexer circuit.

Abstract: In one form, a memory controller includes a command queue and an arbiter. The command queue receives and stores memory access requests. The arbiter includes a plurality of sub-arbiters for providing a corresponding plurality of sub-arbitration winners from among the memory access requests during a controller cycle, and for selecting among the plurality of sub-arbitration winners to provide a plurality of memory commands in a corresponding controller cycle. In another form, a data processing system includes a memory accessing agent for providing memory accesses requests, a memory system, and the memory controller coupled to the memory accessing agent and the memory system.

Abstract: An apparatus and method performs neighborhood-aware virtual to physical address translations. A coalescing opportunity for a first virtual address is determined, based on completing a memory access corresponding to a page walk for a second virtual address. Metadata corresponding to the first virtual address is provided to a page table walk buffer based on the coalescing opportunity and a page walk for the first virtual address is performed based on the metadata corresponding to the first virtual address.

Abstract: Described herein are a method and apparatus for memory vulnerability prediction. A memory vulnerability predictor predicts the reliability of a memory region when it is first accessed, based on past program history. The memory vulnerability predictor uses a table to store reliability predictions and predicts reliability needs of a new memory region. A memory management module uses the reliability information to make decisions, (such as to guide memory placement policies in a heterogeneous memory system).

Abstract: Systems, apparatuses, and methods for routing traffic between clients and system memory are disclosed. A computing system includes system memory and one or more clients, each capable of generating memory access requests. The computing system also includes a communication fabric for transferring traffic between the clients and the system memory. The fabric includes master units for interfacing with clients and grouping write requests with a same target together. The fabric also includes slave units for interfacing with memory controllers and for sending a single write response when each write request in a group has been serviced. When the master unit receives the single write response for the group, it sends a respective acknowledgment response for each of the multiple write requests in the group to clients that generated the multiple write requests.

Abstract: A processing device is provided which includes memory configured to store data and a processor configured to determine, based on convolutional parameters associated with an image, a virtual general matrix-matrix multiplication (GEMM) space of a virtual GEMM space output matrix and generate, in the virtual GEMM space output matrix, a convolution result by matrix multiplying the data corresponding to a virtual GEMM space input matrix with the data corresponding to a virtual GEMM space filter matrix. The processing device also includes convolutional mapping hardware configured to map, based on the convolutional parameters, positions of the virtual GEMM space input matrix to positions of an image space of the image.

Abstract: The described embodiments include an input-output memory management unit (IOMMU) with two or more memory elements and a controller. The controller is configured to select, based on one or more factors, one or more selected memory elements from among the two or more memory elements for performing virtual address to physical address translations in the IOMMU. The controller then performs the virtual address to physical address translations using the one or more selected memory elements.

Abstract: Systems, apparatuses, and methods for implementing a single pass stipple pattern generation process are disclosed. A processor initiates parallel execution of a first and second plurality of wavefronts. A first wavefront of the first plurality of wavefronts converts a first local coordinate into a first global coordinate, wherein the first local coordinate corresponds to a first portion of a primitive. Also, a first wavefront of the second plurality of wavefronts applies a first attribute to the first global coordinate prior to a second wavefront, of the first plurality of wavefronts, converting a second local coordinate of a second portion of the primitive into a second global coordinate. The second plurality of wavefronts generate image data based on applying the first attribute to global coordinates generated by the first plurality of wavefronts, and the image data is conveyed for display on a display device.

Abstract: Systems, apparatuses, and methods for dynamically adjusting data bandwidth to fit a predicted available bandwidth of a link are disclosed. A system includes a transmitter and a receiver communicating wirelessly over a wireless link. The system monitors a status of the wireless link at a plurality of points in time. The system trains a predictive model using a plurality of indicators of the status of the link at the plurality of points in time. The model generates a prediction of the available bandwidth of the link based on the plurality of indicators. Then, the transmitter dynamically adjusts an amount of data sent on the link to fit the predicted available bandwidth of the link.

Abstract: A method of message-based communication is provided which includes executing, on one or more accelerated processing units, a plurality of groups of work items, receiving a first message from a first group of work items of the plurality of groups of work items executing on the one or more accelerated processing units and storing the first message at a first segment of memory allocated to a second group of work items of the plurality of groups of work items executing on the accelerated processing unit.

Abstract: A generator for generating artificial data, and training for the same. Data corresponding to a first label is altered within a reference labeled data set. A discriminator is trained based on the reference labeled data set to create a selectively poisoned discriminator. A generator is trained based on the selectively poisoned discriminator to create a selectively poisoned generator. The selectively poisoned generator is tested for the first label and tested for the second label to determine whether the generator is sufficiently poisoned for the first label and sufficiently accurate for the second label. If it is not, the generator is retrained based on the data set including the further altered data. The generator includes a first ANN to input first information and output a set of artificial data that is classifiable using a first label and not classifiable using a second label of the set of labeled data.

Abstract: A method and apparatus for physical layer bypass data transmission between physical coding sub-layers (PCS) includes encoding the data for transmission over a serial low-speed link. The data is transmitted from a first PCS via a serial connection over a serializer/deserializer (SERDES) transmission bypass path The data is received by a second PCS via a SERDES receive bypass path.

Abstract: Various computing network messaging techniques and apparatus are disclosed. In one aspect, a method of computing is provided that includes executing a first thread and a second thread. A message is sent from the first thread to the second thread. The message includes a domain descriptor that identifies a first location of the first thread and a second location of the second thread.

Abstract: A prefetcher maintains the state of stored prefetch information, such as a prefetch confidence level, when a prefetch would cross a memory page boundary. The maintained prefetch information can be used both to identify whether the stride pattern for a particular sequence of demand requests persists after the memory page boundary has been crossed, and to continue to issue prefetch requests according to the identified pattern. The prefetcher therefore does not have re-identify a stride pattern each time a page boundary is crossed by a sequence of demand requests, thereby improving the efficiency and accuracy of the prefetcher.