Abstract: A technique for performing ray tracing operations is provided. The technique includes, testing a plurality of bounding boxes for intersection with a ray in parallel, wherein the plurality of bounding boxes are specified by a plurality of box data items of a parent box node of a bounding volume hierarchy; determining that, for a first child node that is pointed to by a two or more node pointers specified by two or more box data items of the plurality of box data items, at least one bounding box specified by the two or more box data items is intersected by the ray; and in response to the determining, traversing to the first child node.

Abstract: The disclosed computer-implemented method can include reaching, by a chiplet involved in carrying out an operation for a process, a synchronization barrier. The method can additionally include receiving, by the chiplet, dedicated control messages pushed to the chiplet by other chiplets involved in carrying out the operation for the process, wherein the dedicated control messages are pushed over a control network by the other chiplets. The method can also include advancing, by the chiplet, the synchronization barrier in response to receipt of the dedicated control messages. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The disclosed device includes a collectives engine that can offload collectives communications of multiple nodes and perform collective operations. The collectives engine can manage a direct mapping scheme of local memories of the nodes for access by the collectives engine. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A technique for performing ray tracing operations is provided. The technique includes, in a first iteration of a ray traversal technique, traversing to an instance node of a bounding volume hierarchy; in a second iteration of the ray traversal technique that is subsequent to the first iteration, transforming a ray based on an instance transform of the instance node to generate a transformed ray; and in the second iteration, performing a ray-box intersection test for box node data of the instance node based on the transformed ray.

Abstract: A processing device and a method of auto-tiled workload processing is provided. The processing device includes memory and a processor. The processor is configured to store instructions for operations to be executed on an image to be divided into a plurality of tiles, store information associated with the operations, select one of the operations for execution and execute an auto-tiling plan for the operation based on the information associated with the operations. The auto-tiling plan comprises, for example, determining a number of tiles used to divide the image and determining a size of one or more of the tiles of the image.

Abstract: A graphics processing unit (GPU) of a processing system transmits pixel data for a frame to a display in a compressed burst, so that the pixel data is communicated at a rate that is higher than the rate at which the display scans out the pixel data to refresh the frame at a display panel. By transmitting pixel data for the frame in a compressed burst, the GPU shortens the time spent transmitting the pixel data and extends the time before the next frame of pixel data is to be transmitted. During the extended time before the next frame of pixel data is to be transmitted, the GPU saves power by placing portions of the processing system in a reduced power mode.

Abstract: A technique for operating an auxiliary processing device is provided. The technique includes based on a first request specifying a handle received from a client, requesting work be performed via a first auxiliary processing device mapped to the handle; responsive to restoration from hibernation, updating a mapping for the handle to refer to a second auxiliary processing device; and based on a second request specifying the handle received from the client, requesting work be performed via the second auxiliary processing device.

Abstract: A read clock circuit selectively provides a read clock signal from a memory to a memory controller over a memory bus. A pulse-amplitude modulation (PAM) driver including an input and an output capable of driving at least three levels indicating respective digital values. A digital control circuit is coupled to the PAM driver and operable to cause the PAM driver to provide a preamble signal before the read clock signal, the preamble signal including an initial toggling state in which the PAM driver toggles between two selected levels at a first rate, and a final toggling state in which the PAM driver toggles between two selected levels at a second rate higher than the first rate, with a length of the initial toggling state and a length of the final toggling state are based on values in a mode register.

Abstract: A system and method for texture decompression is described. The method comprises receiving a compressed texture block including two or more disjoint subsets of data and decompressing the compressed texture block. The decompressing includes decompressing each of the two or more disjoint subsets in the compressed texture block to form texels. The two or more disjoint subsets include a first disjoint subset having a first set of color endpoints and a first index value for a first texel, and a second disjoint subset having a second set of color endpoints.

Abstract: A method for hardware management of DMA transfer commands includes accessing, by a first DMA engine, a DMA transfer command and determining a first portion of a data transfer requested by the DMA transfer command. Transfer of a first portion of the data transfer by the first DMA engine is initiated based at least in part on the DMA transfer command. Similarly, a second portion of the data transfer by a second DMA engine is initiated based at least in part on the DMA transfer command. After transferring the first portion and the second portion of the data transfer, an indication is generated that signals completion of the data transfer requested by the DMA transfer command.

Abstract: The disclosed computer-implemented method for generating remedy recommendations for power and performance issues within semiconductor software and hardware. For example, the disclosed systems and methods can apply a rule-based model to telemetry data to generate rule-based root-cause outputs as well as telemetry-based unknown outputs. The disclosed systems and methods can further apply a root-cause machine learning model to the telemetry-based unknown outputs to analyze deep and complex failure patterns with the telemetry-based unknown outputs to ultimately generate one or more root-cause remedy recommendations that are specific to the identified failure and the client computing device that is experiencing that failure.

Abstract: Apparatuses, systems and methods for routing requests and responses targeting a shared resource. A queue in a communication fabric is located in a path between the requesters and a shared resource. In some embodiments, the shared resource is a shared address translation cache stored in an endpoint. The physical channel between the queue and the shared resource supports multiple virtual channels. The queue assigns at least one entry to each virtual channel of a group of virtual channels where the group includes a virtual channel for each address translation request type from a single requester of the multiple requesters. When the at least one entry for a given requester is de-allocated, the queue allocates this entry only with requests from the assigned virtual channel even if the empty entry is the only available entry of the queue.

Abstract: Systems, apparatuses, and methods for converting pixel data to a custom swizzle mode are disclosed. A graphics engine receives data in a pre-defined swizzle mode. The graphics engine determines a custom swizzle mode for the data that has directionality aligned to the data itself to further optimize deltas that are used for compressing the data. The graphics engine groups incoming data into group of two neighboring pixels in both the horizontal and vertical directions. The graphics engine scores horizontal and vertical groupings against each other to make a first swizzle mode bit selection. Then the graphics engine increases the grouping of pixels to include additional pixels and scores the increased groupings against each other to make subsequent swizzle mode bit selections. The data is reswizzled into the custom swizzle mode and provided to a compressor to be compressed.

Abstract: A technique for encoding video is provided. The technique includes for a first portion of a first frame that is encoded by a first encoder in parallel with a second portion of the first frame that is encoded by a second encoder, determining a historical complexity distribution; determining a first bit budget for the first portion of the first frame based on the historical complexity distribution; and encoding the first portion of the first frame by the first encoder, based on the first bit budget.

Abstract: The disclosed computer-implemented method for configuring fan speeds can include (i) measuring an air temperature at the air intake of a fan that cools a hardware processing unit of a computing device, (ii) adjusting a rotational speed for the fan based on the air temperature at the air intake of the fan and at least one additional parameter measured around the time of measuring the temperature of the air, and (iii) sending, to the fan, an instruction to rotate at the rotational speed. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A compilation technique is provided. The technique includes including a first instruction into a first executable for a first auxiliary processor, wherein the first instruction specifies execution by the first auxiliary processor; and including a second instruction into the first executable, wherein the second instruction targets resources that have affinity with the first auxiliary processor.

Abstract: Systems, apparatuses, and methods for implementing a safety monitor framework for a safety-critical computer vision (CV) application are disclosed. A system includes a safety-critical CV application, a safety monitor, and a CV accelerator engine. The safety monitor receives an input image, test data, and a CV graph from the safety-critical CV application. The safety monitor generates a modified image by adding additional objects outside of the input image. The safety monitor provides the modified image and CV graph to the CV accelerator which processes the modified image and provides outputs to the safety monitor. The safety monitor determines the likelihood of erroneous processing of the original input image by comparing the outputs for the additional objects with a known good result. The safety monitor complements the overall fault coverage of the CV accelerator engine and covers faults only observable at the level of the CV graph.

Abstract: A method, computer system, and a non-transitory computer-readable storage medium for performing primitive batch binning are disclosed. The method, computer system, and non-transitory computer-readable storage medium include techniques for generating a primitive batch from a plurality of primitives, computing respective bin intercepts for each of the plurality of primitives in the primitive batch, and shading the primitive batch by iteratively processing each of the respective bin intercepts computed until all of the respective bin intercepts are processed.

Abstract: Techniques described herein provide users with the ability to persistently adjust settings for boot-time features (BTF) of a computing device. A user requests a particular BTF configuration adjustment for a device via a device driver. The driver instructs trusted firmware of the device to store a boot override record in persistent storage accessible by a bootloader for the device. Upon implementation of the boot sequence for the device, the bootloader applies the changes reflected in the record to BTF configuration data. The boot override information is persistently available to the bootloader, which ensures that the configuration changes that the boot override record(s) represent are applied to the BTFs of the device until the boot override record(s) are cleared or invalidated. Further, to ensure the security of boot override record(s), the trusted firmware generates, for each record, an HMAC tag using an HMAC key derived from a Chip Endorsement Fused Secret from the hardware.

Abstract: Systems, apparatuses, and methods for implementing a unified kernel virtual address space for heterogeneous computing are disclosed. A system includes at least a first subsystem running a first kernel, an input/output memory management unit (IOMMU), and a second subsystem running a second kernel. In order to share a memory buffer between the two subsystems, the first subsystem allocates a block of memory in part of the system memory controlled by the first subsystem. A first mapping is created from a first logical address of the kernel address space of the first subsystem to the block of memory. Then, the IOMMU creates a second mapping to map the physical address of that block of memory from a second logical address of the kernel address space of the second subsystem. These mappings allow the first and second subsystems to share buffer pointers which reference the block of memory.