Abstract: Embodiments herein provide for improved store-to-load-forwarding (STLF) logic and linear aliasing effect reduction logic. In one embodiment, a load instruction to be executed is selected. Whether a first linear address associated with said load instruction matches a linear address of a store instruction of a plurality of store instructions in a queue is determined. Data associated with said store instruction for executing said load instruction is forwarded, in response to determining that the first linear address matches the linear address of the store instruction.

Abstract: Shader resources may be specified for input to a shader using a hierarchical data structure which may be referred to as a descriptor set. The descriptor set may be bound to a bind point of the shader and may contain slots with pointers to memory containing shader resources. The shader may reference a particular slot of the descriptor set using an offset, and may change shader resources by referencing a different slot of the descriptor set or by binding or rebinding a new descriptor set. A graphics pipeline may be specified by creating a pipeline object which specifies a shader and a rendering context object, and linking the pipeline object. Part or all of the pipeline may be validated, cross-validated, or optimized during linking.

Abstract: Briefly, methods and apparatus to rebalance workloads among processing cores utilizing a hybrid work donation and work stealing technique are disclosed that improve workload imbalances within processing devices such as, for example, GPUs. In one example, the methods and apparatus allow for workload distribution between a first processing core and a second processing core by providing queue elements from one or more workgroup queues associated with workgroups executing on the first processing core to a first donation queue that may also be associated with the workgroups executing on the first processing core. The method and apparatus also determine if a queue level of the first donation queue is beyond a threshold, and if so, steal one or more queue elements from a second donation queue associated with workgroups executing on the second processing core.

Abstract: A system and method embodying some aspects for communicating between nodes in a network-on-chip are provided. The system comprises a microprocessing chip and a plurality of connection paths. The microprocessing chip comprises sixteen processing nodes disposed on the chip. The plurality of connection paths are configured such that each is at most three hops away front any other node. Each node also has connection paths to at most three other nodes.

Abstract: A circuit for detecting out-of-band signals is disclosed. In one embodiment, the circuit includes a first differential circuit configured to level shift and positively rectify a differential input signal to produce a first output component of a differential output signal. The detector further includes a second differential circuit configured to level shift and negatively rectify the differential input signal to produce a second output component of the differential output signal. A third differential circuit is configured to level shift and output first and second fixed voltages based on an input reference voltage and a ground voltage. The circuit is configured to provide the differential output signal and the first and second fixed voltages to an indicator circuit configured to assert an indication responsive to detecting that a differential voltage of the differential output signal is greater than a differential voltage of the first and second fixed voltages.

Abstract: Various method and apparatus embodiments for training a delay for enabling a data strobe signal in a memory subsystem are disclosed. In one embodiment, a system includes a memory controller configured to receive a data strobe signal. The memory controller includes a training circuit. The training circuit includes a first storage circuit coupled to receive the data strobe signal on a data input and an enable signal on a clock input, and a training unit configured to, based on an output signal received from the first flip-flop, adjust a phase of the enable signal until an assertion of the enable signal coincides with a preamble indication in the data strobe signal.

Abstract: Some embodiments include a partitioning mechanism that partitions a cache memory into sub-partitions for sub-entities. In the described embodiments, the cache memory is initially partitioned into two or more partitions for one or more corresponding entities. During a partitioning operation, the partitioning mechanism is configured to partition one or more of the partitions in the cache memory into two or more sub-partitions for one or more sub-entities of a corresponding entity. A cache controller then uses a corresponding sub-partition for memory accesses by the one or more sub-entities.

Abstract: Some embodiments include a computing device with a control circuit that handles memory requests. The control circuit checks one or more conditions to determine when a memory request should be bypassed to a main memory instead of sending the memory request to a cache memory. When the memory request should be bypassed to a main memory, the control circuit sends the memory request to the main memory. Otherwise, the control circuit sends the memory request to the cache memory.

Abstract: The described embodiments include a cache controller that configures a cache management mechanism. In the described embodiments, the cache controller is configured to monitor at least one structure associated with a cache to determine at least one cache block that may be accessed during a future access in the cache. Based on the determination of the at least one cache block that may be accessed during a future access in the cache, the cache controller configures the cache management mechanism.

Abstract: The described embodiments include a computing device that handles memory requests. In some embodiments, when a memory request is to be sent to a cache in the computing device or to be bypassed to a next lower level of a memory hierarchy in the computing device based on expected memory request resolution times, a bypass mechanism is configured to send the memory request to the cache or bypass the memory request to the next lower level of the memory hierarchy.

Abstract: A method and apparatus for powering up an integrated circuit (IC). An IC includes a plurality of power domains each coupled to receive power from one of a plurality of power sources. Each power domain includes a power-sensing unit. A power-sensing unit in a first one of the plurality of power domains is coupled to receive a first power ok signal from an upstream power domain, and is configured to assert a second power ok signal to be provided to a second power domain. A power-sensing unit in the second power domain is coupled to detect the presence of voltage in the first power domain, and to receive the first power ok signal. When the power-sensing unit in the second power domain has both sensed the presence of power in the first power domain and received the second power ok signal, a third power ok signal is asserted.

Abstract: A semiconductor device fabrication process includes forming insulating mandrels over replacement metal gates on a semiconductor substrate with first gates having sources and drains and at least one second gate being isolated from the first gates. Mandrel spacers are formed around each insulating mandrel. The mandrels and mandrel spacers include the first insulating material. A second insulating layer of the second insulating material is formed over the transistor. One or more first trenches are formed to the sources and drains of the first gates by removing the second insulating material between the insulating mandrels. A second trench is formed to the second gate by removing portions of the first and second insulating materials above the second gate. The first trenches and the second trench are filled with conductive material to form first contacts to the sources and drains of the first gates and a second contact to the second gate.

Abstract: The present disclosure relates to a method and apparatus for dynamically controlling power consumption by at least one processor. A power management method includes monitoring, by power control logic of the at least one processor, performance data associated with each of a plurality of executions of a repetitive workload by the at least one processor. The method includes adjusting, by the power control logic following an execution of the repetitive workload, an operating frequency of at least one of a compute unit and a memory controller upon a determination that the at least one processor is at least one of compute-bound and memory-bound based on monitored performance data associated with the execution of the repetitive workload.

Abstract: A method for power management is disclosed. The method may include monitoring requests for access to a memory of a memory subsystem by one or more processor cores; and monitoring requests for access to the memory conveyed by an input/output (I/O) unit. The method may further include determining if at least a first amount of time has elapsed since any one of the processor cores has asserted a memory access request and determining if at least a second amount of time has elapsed since the I/O unit has conveyed a memory access request. A first signal may be asserted if the first and second amounts of time have elapsed. A memory subsystem may be transitioned from operating in a full power state to a first low power state responsive to assertion of the first signal.

Abstract: A decimal floating-point (DFP) adder includes a decimal leading-zero anticipator (LZA). The DFP adder receives DFP operands. Each operand includes a significand, an exponent, a sign bit and a leading zero count for the significand. The DFP adder adds or subtracts the DFP operands to obtain a DFP result. The LZA determines the leading zero count associated with the significand of the DFP result. The LZA operates at least partially in parallel with circuitry (in the DFP adder) that computes the DFP result. The LZA does not wait for that circuitry to finish computation of the DFP result. Instead it “anticipates” the number of leading zeros that the result's significand will contain.

Abstract: A hybrid prefetch method and apparatus is disclosed. A processor includes a hybrid prefetch unit configured to generate addresses for accessing data from a system memory. The hybrid prefetch unit includes a first prediction unit configured to generate a first memory address according to a first prefetch algorithm and a second prediction unit configured to generate a second memory address according to a second prefetch algorithm. The hybrid prefetcher further includes an arbitration unit configured to select one of the first and second memory addresses and further configured to provide the selected one of the first and second memory addresses during a prefetch operation.

Abstract: A semiconductor device fabrication process includes forming a gate of a transistor on a semiconductor substrate using a hard mask. The hard mask is selectively removed in one or more selected regions over the gate. The removal of the hard mask in the selected regions allows the gate to be connected to an upper metal layer through at least one insulating layer located substantially over the transistor. Conductive material is deposited in one or more trenches formed through the at least one insulating layer. The conductive material forms a local interconnect to the gate in at least one of the selected regions.

Abstract: A semiconductor device fabrication process includes forming insulating mandrels over one or more replacement metal gates on a semiconductor substrate. The mandrels include a first insulating material. Each mandrel has approximately the same width as its underlying gate with each mandrel being at least as wide as its underlying gate. Mandrel spacers are formed around each insulating mandrel. The mandrel spacers include the first insulating material. Each mandrel spacer has a profile that slopes from being wider at the bottom to narrower at the top. A second insulating layer of the second insulating material is formed over the transistor. Trenches to the sources and drains of the gates are formed by removing the second insulating material from portions of the transistor between the mandrels. Trench contacts to the sources and drains of the gates are formed by depositing conductive material in the first trenches.

Abstract: An electrically programmable reticle is made using at least one electrochromatic layer that changes its optical transmissibility in response to applied voltages. Transparent conductor layers are configured to the desired patterns. The electrically programmable reticles are either patterned in continuous forms that have separately applied voltages or in a matrix of rows and columns that are addressed by row and column selects such that desired patterns are formed with the application of a first voltage level and reset with the application of a second voltage level.

Abstract: A circuit for detecting out-of-band signals is disclosed. In one embodiment, the circuit includes a first differential circuit configured to level shift and positively rectify a differential input signal to produce a first output component of a differential output signal. The first differential circuit is further configured to generate and provide a common mode voltage of the differential input signal as a second component of the differential output signal. The circuit further includes a second differential circuit configured to level shift and output first and second fixed voltages based on an input reference voltage and a ground voltage. The circuit is configured to provide the differential output signal and the first and second fixed voltages to an indicator circuit configured to assert an indication responsive to detecting that a differential voltage of the differential output signal is greater than a differential voltage of the first and second fixed voltages.