Abstract: A system and method for virtually addressing an array of accelerator tiles of a mixed-signal integrated circuit includes testing each of a plurality of distinct matrix multiply accelerator (MMA) tiles of a grid of MMA tiles, the grid of MMA tiles being defined by the plurality of distinct grid of MMA tiles being arranged in a plurality of rows and a plurality of columns along an integrated circuit, each of the plurality of distinct MMA tiles within the grid of MMA tiles having a distinct physical address on the integrated circuit; identifying one or more defective MMA tiles within the grid of MMA tiles based on the testing; and configuring the grid of MMA tiles with a plurality of virtual addresses for routing data to or routing data from one or more non-defective MMA tiles of grid of MMA tiles based on identifying the one or more defective MMA tiles.

Abstract: A system and method for virtually addressing an array of accelerator tiles of a mixed-signal integrated circuit includes testing each of a plurality of distinct matrix multiply accelerator (MMA) tiles of a grid of MMA tiles, the grid of MMA tiles being defined by the plurality of distinct grid of MMA tiles being arranged in a plurality of rows and a plurality of columns along an integrated circuit, each of the plurality of distinct MMA tiles within the grid of MMA tiles having a distinct physical address on the integrated circuit; identifying one or more defective MMA tiles within the grid of MMA tiles based on the testing; and configuring the grid of MMA tiles with a plurality of virtual addresses for routing data to or routing data from one or more non-defective MMA tiles of grid of MMA tiles based on identifying the one or more defective MMA tiles.

Abstract: A system and method for virtually addressing an array of accelerator tiles of a mixed-signal integrated circuit includes testing each of a plurality of distinct matrix multiply accelerator (MMA) tiles of a grid of MMA tiles, the grid of MMA tiles being defined by the plurality of distinct grid of MMA tiles being arranged in a plurality of rows and a plurality of columns along an integrated circuit, each of the plurality of distinct MMA tiles within the grid of MMA tiles having a distinct physical address on the integrated circuit; identifying one or more defective MMA tiles within the grid of MMA tiles based on the testing; and configuring the grid of MMA tiles with a plurality of virtual addresses for routing data to or routing data from one or more non-defective MMA tiles of grid of MMA tiles based on identifying the one or more defective MMA tiles.

Abstract: A system and method for virtually addressing an array of accelerator tiles of a mixed-signal integrated circuit includes testing each of a plurality of distinct matrix multiply accelerator (MMA) tiles of a grid of MMA tiles, the grid of MMA tiles being defined by the plurality of distinct grid of MMA tiles being arranged in a plurality of rows and a plurality of columns along an integrated circuit, each of the plurality of distinct MMA tiles within the grid of MMA tiles having a distinct physical address on the integrated circuit; identifying one or more defective MMA tiles within the grid of MMA tiles based on the testing; and configuring the grid of MMA tiles with a plurality of virtual addresses for routing data to or routing data from one or more non-defective MMA tiles of grid of MMA tiles based on identifying the one or more defective MMA tiles.

Abstract: A SoC interconnect network topology is represented. The corresponding SoC floorplan is divided into windows, which are contiguous and non-overlapping. Within each window a subnetwork of the SoC interconnect network topology is defined that includes links or communication paths between IP blocks in the window as well as links or communication paths that traverse the window. At the shared boundaries of the windows, ports are added and defined as virtual ports. The overall SoC topology can be optimized and synthesized by optimizing each window independently and then incrementally optimizing all links, from end-to-end, that traverse two or more windows. The SoC topology is edited, such as by adding switches, removing switches, and adding and removing switches on routes. An initial location of elements within the floorplan is automatically computed and recommended. Locations can also be edited. Statistical metrics are calculated, including wire length, switch area, SoC area, and maximum signal propagation rate.

Abstract: A method of instruction issue (3200) in a microprocessor (1100, 1400, or 1500) with execution pipestages (E1, E2, etc.) and that executes a producer instruction Ip and issues a candidate instruction I0 (3245) having a source operand dependency on a destination operand of instruction Ip. The method includes issuing the candidate instruction I0 as a function (1720, 1950, 1958, 3235) of a pipestage EN(I0) of first need by the candidate instruction for the source operand, a pipestage EA(Ip) of first availability of the destination operand from the producer instruction, and the one execution pipestage E(Ip) currently associated with the producer instruction. A method of data forwarding (3300) in a microprocessor (1100, 1400, or 1500) having a pipeline (1640) having pipestages (E1, E2, etc.

Abstract: A power management method for a pipelined computer system in accordance with one or both of a power management signal and a power management instruction.

Abstract: A pipelined computer system with power management control in accordance with one or both of a power management signal and a power management instruction.

Abstract: An instruction-initiated method for suspending operation of a pipelined data processor by selectively disabling a clock signal to pipeline subcircuitry in response to an instruction executed by the pipeline subcircuitry.

Abstract: A processor comprises decode logic that determines an instruction type for each instruction fetched, a first level cache, a second level cache coupled to the first level cache, and control logic operatively coupled to the first and second level caches. The control logic preferably causes cache linefills to be performed to the first level cache upon cache misses for a first type of instruction, but precludes linefills from being performed to the first level cache for a second type of instruction.

Abstract: An instruction-initiated power management method for a pipelined data processor by which a clock signal to pipeline subcircuitry is selectively disabled in response to an instruction executed by the pipeline subcircuitry.

Abstract: A signal-initiated method for suspending operation of a pipelined data processor by selectively disabling a clock signal to pipeline subcircuitry in response to at least one control signal.

Abstract: A signal-initiated power management method for a pipelined data processor by which a clock signal to pipeline subcircuitry is selectively disabled in response to at least one control signal.

Abstract: A pipelined data processor with instruction-initiated power management control in which a plurality of subcircuits, including pipeline subcircuitry and circuitry for generating and controlling at least one clock signal are responsive to an instruction executed by the pipeline subcircuitry by selectively disabling a clock signal to the pipeline subcircuitry.

Abstract: A pipelined data processor with signal-initiated power management control in which a plurality of subcircuits, including pipeline subcircuitry, and circuitry for generating and controlling at least one clock signal are responsive to at least one control signal by disabling a clock signal to the pipeline subcircuitry.

Abstract: In accordance with the presently claimed invention, power consumption reduction control is provided to a processor used to execute instructions for data processing. A power management control signal is provided to the processor in accordance with conditions associated with the processor being operated in normal and reduced power consumption modes of operation, and an acknowledgement signal indicative of such reduced power consumption mode of operation is returned in correspondence with the power management control signal.

Abstract: Power consumption reduction control circuitry external and coupled to a processor used to execute instructions for data processing. A power management control signal is provided to the processor in accordance with conditions associated with the processor being operated in normal and reduced power consumption modes of operation, and an acknowledgement signal indicative of such reduced power consumption mode of operation is returned in correspondence with the power management control signal.

Abstract: A technique for invoking a low power operational mode in response to a halt instruction is used in a computer system that includes a processor coupled to external logic. The processor includes at least (i) a pipeline subcircuit to execute programmed instructions, including halt instructions, (ii) an interrupt handling subcircuit to handle interrupts generated by external interrupt logic, and (iii) clock generator circuitry that supplies clock signals to the pipeline and interrupt handling subcircuits. In response to execution of a halt instruction, the processor (i) enters the low power operational mode in which power consumption is reduced at least for the pipeline subcircuit, but without stopping the supply of clock signals to the interrupt handling subcircuit, and (ii) generates an acknowledgement signal to the external logic indicating that the clock signals to the pipeline subcircuit are being stopped, thereby entering the low power operational mode.

Abstract: A superscalar superpipelined microprocessor having a write buffer located between the core and cache is disclosed. The write buffer is controlled to store the results of write operations to memory until such time as the cache becomes available, such as when no high-priority reads are to be performed. The write buffer includes multiple entries that are split into two circular buffer sections for facilitating the interaction with the two pipelines of the core; cross-dependency tables are provided for each write buffer entry to ensure that the data is written from the write buffer to memory in program order, considering the possibility of prior data present in the opposite section. Non-cacheable reads form memory are also ordered in program order with the writing of data from the write buffer. Features for handling speculative execution, detecting and handling data dependencies and exceptions, and performing special write functions (misaligned writes and gathered writes) are also disclosed.

Abstract: A microprocessor comprises a plurality of instruction pipelines having a plurality of stages for processing a stream of instructions, circuitry for simultaneously issuing instructions into two or more of the pipelines without regard to whether one of the simultaneously issued instructions has a data dependency on other of the simultaneously issued instructions, detecting circuitry for detecting dependencies between instructions in the pipelines and circuitry for controlling the flow of instructions through the pipelines such that an instruction is not delayed due to a data dependency on another instruction unless the data dependency must be resolved for proper processing of the instruction in its current stage.