Abstract: A multi-die integrated circuit (IC) can include an interposer and a first die coupled to the interposer. The first die can include a data processing engine (DPE) array, wherein the DPE array includes a plurality of DPEs and a DPE interface coupled to the plurality of DPEs. The DPE interface has a logical interface and a physical interface. The multi-die IC also can include a second die coupled to the interposer. The second die can include a die interface. The DPE interface and the die interface are configured to communicate through the interposer.

Abstract: A switch with clock-gating control and a method for clock gating a switch are described herein. In one example, the method generally includes detecting a state of one or more input ports and a state of one or more output ports of the switch, determining whether the state of the one or more input ports and the state of the one or more output ports has been stable for a preset number of clock cycles, and gating the switch from a clock signal until the state of the one or more input ports or the state of the one or more output ports change upon determining the states have been stable for the preset number of the cycles.

Abstract: A device may include a plurality of data processing engines. Each of the data processing engines may include a core and a memory module. The plurality of data processing engines may be organized in a plurality of rows. Each core may be configured to communicate with other neighboring data processing engines of the plurality of data processing engines by shared access to the memory modules of the neighboring data processing engines.

Abstract: An example core for a data processing engine (DPE) includes a register file, a processor, coupled to the register file. The processor includes a multiply-accumulate (MAC) circuit, and permute circuitry coupled between the register file and the MAC circuit, the permute circuitry configured to concatenate at least one pair of outputs of the register file to provide at least one input to the MAC circuit. The core further includes an instruction decoder, coupled to the processor, configured to decode a very large instruction word (VLIW) to set a plurality of parameters of the processor, the plurality of parameters including first parameters of the permute circuitry and second parameters of the MAC circuit.

Abstract: A device may include a plurality of data processing engines, a subsystem, and an SoC interface block coupled to the plurality of data processing engines and the subsystem. The SoC interface block may be configured to exchange data between the subsystem and the plurality of data processing engines.

Abstract: A device may include a plurality of data processing engines. Each of the data processing engines may include a core and a memory module. The plurality of data processing engines may be organized in a plurality of rows. Each core may be configured to communicate with other neighboring data processing engines of the plurality of data processing engines by shared access to the memory modules of the neighboring data processing engines.

Abstract: A device may include a plurality of data processing engines, a subsystem, and an SoC interface block coupled to the plurality of data processing engines and the subsystem. The SoC interface block may be configured to exchange data between the subsystem and the plurality of data processing engines.

Abstract: A method and apparatus for adaptively tuning an integrated circuit are disclosed. For example, an integrated circuit (IC) comprises a monitored path comprising circuit elements operating on a clock signal, where a last circuit element of the circuit elements comprises a first flip flop. The IC also comprises a second flip flop operating on an early clock signal, where the early clock signal is phase shifted from the clock signal, and where the second flip flop is coupled to the monitored path prior to the last circuit element. The IC also comprises a transition detection module for detecting when an output from the first flip flop toggles, and an error prediction module to detect a potential error on the monitored path. The IC comprises a controller that is configured to scale a voltage or a frequency of the IC.

Abstract: A method for asynchronous time multiplexing of information with synchronous interfacing includes, responsive to a first edge of a clock signal, asynchronously loading first data, including first multiple sets of data for multiple operations, into a first asynchronous shift register. The first data is asynchronously unloaded from the first asynchronous shift register to a function block for processing to provide second data, including second multiple sets of data as results of the multiple operations. The second data is asynchronously loaded into a second asynchronous shift register. Responsive to a second edge of the clock signal, the second data is asynchronously unloaded from the second asynchronous shift register as the results of the multiple operations. The first edge and the second edge of the clock signal are associated with a same period of the clock signal.

Abstract: A method for determining or configuring supply voltage and threshold voltage for a design implementation of a given electronic design, includes: determining a first set of supply voltage-threshold voltage combinations that meet timing requirements for the design implementation; performing power analysis using a processor; and selecting a supply voltage-threshold voltage combination from the first set of supply voltage-threshold voltage combinations based at least in part on a result from the power analysis, wherein the selected supply voltage-threshold voltage combination provides an optimal amount of power consumption for the design implementation.

Abstract: A method and circuit for device specific configuration of an operating voltage is provided. A circuit design is analyzed to determine a maximum gate-level delay for the circuit design. A minimum voltage value corresponding to the maximum gate-level delay is determined along with a default voltage value corresponding to a default gate-level delay. A voltage scaling factor corresponding to the minimum voltage and default voltage values is determined. The circuit design is synthesized such that the synthesized design includes the voltage scaling value. The synthesized design specifies setting an operating voltage to a value of a startup voltage value scaled by the voltage scaling value.

Abstract: Approaches for generating a hardware definition from a program specified in a high-level language. In one approach, a first set of blocks of instructions in the high-level language program is identified. Each block in the first set is bounded by a respective loop designation in the high-level language. For each block in the first set, an associated respective second set of one or more blocks of the program is identified. Each block in the second set is outside the block in the first set. A hardware definition of the program is generated and stored. For each block in the first set, the hardware definition specifies power-reducing circuitry for one or more blocks in the associated second set. The power-reducing circuitry is controlled based on a status indication from the hardware definition of the block in the first set.

Abstract: In a memory device, an array of memory cells is coupled between a virtual ground node and a supply node. First and second transistors are coupled in source-drain parallel between the virtual ground node and a ground bus. The first transistor is substantially larger than the second transistor. A control circuit provides a first gate signal to a gate of the first transistor and a second gate signal to a gate of the second transistor. The control circuit includes: a configuration memory cell providing a first control signal; an interconnect providing a second control signal; and control logic receiving the first and second control signals and providing the first gate signal. The array of memory cells has three modes responsive to the first and second gate signals, where the three modes include an active mode, a first sleep mode, and a second sleep mode.

Abstract: A programmable integrated circuit having a plurality of individually controlled voltage domains. Each voltage domain includes logic circuitry powered by a respective power network. The voltage magnitude of each power network is independently selectable. Each of a plurality of level shifters couples a first and second one of the voltage domains, couples a first port of the logic circuitry of the first voltage domain to a second port of the logic circuitry of the second voltage domain, and shifts from a first signaling protocol of the first port to a second signaling protocol of the second port. The first signaling protocol is referenced to the voltage magnitude of the first voltage domain, and the second signaling protocol is referenced to the voltage magnitude of the second voltage domain. Means are disclosed for controlling the voltage magnitude of the respective power network of one or more of the voltage domains.

Abstract: Methods of enabling functions of a design to be implemented in an integrated circuit device are disclosed. An exemplary method comprises applying test data to a plurality of dice having different element types for implementing circuits, wherein the plurality of dice have a common layout of the different element types for implementing the circuits; receiving output data from the plurality of dice in response to applying the test data to the plurality of dice; analyzing the output data from the plurality of dice; transforming by a computer the output data to characterization data comprising timing data associated with the different element types for implementing circuits, wherein the characterization data comprises data associated with regions of the dice, and storing the characterization data. A computer program product for enabling functions of a design to be implemented in an integrated circuit device is also disclosed.

Abstract: A method for optimizing a high-level circuit architecture for an integrated circuit is described. Descriptions of components of the circuit architecture and optimization goals for the components are received. At least one stopping criterion for the cost functions is received. Implementations for the components are iteratively generated to provide a system from a combination of the implementations. The implementations of the components are iteratively optimized until the at least one stopping criterion is satisfied. The optimizing includes obtaining estimation models for determining cost estimates for the implementations and iteratively optimizing the implementations responsive to the cost estimates.

Abstract: An apparatus and method for the dynamic detection and compensation of performance variations within an integrated circuit (IC) is provided to detect performance variations within the IC at any stage of test or operation. An arbitrary reference signal is utilized in conjunction with an internal oscillation device to establish a speed reference that may be used to characterize the IC. Dynamic detection and compensation may also be configured within a plurality of geographic locations within the IC, so that performance variations may be detected and compensated. Test data that is indicative of the IC's performance may be dynamically generated continuously, or at programmable intervals, so that performance variations caused by virtually any source may be substantially detected and compensated at any point in time of the IC's life cycle.

Abstract: A method of operating an integrated circuit (“IC”) is disclosed. The method includes identifying one or more unused or inactive resources of the IC which will not be used in a circuit design or which are inactive during operation of the IC. The method also includes enabling resources of the IC which will be used in the circuit design, and also disabling one or more unused or inactive resources of the IC from one or more power supply terminals in response to configuration values which are stored in memory cells.

Abstract: A method and circuit for device specific configuration of an operating voltage is provided. A circuit design is analyzed to determine a maximum gate-level delay for the circuit design. A minimum voltage value corresponding to the maximum gate-level delay is determined along with a default voltage value corresponding to a default gate-level delay. A voltage scaling factor corresponding to the minimum voltage and default voltage values is determined. The circuit design is synthesized such that the synthesized design includes the voltage scaling value. The synthesized design specifies setting an operating voltage to a value of a startup voltage value scaled by the voltage scaling value.

Abstract: Power management with a packaged multi-die integrated circuit (IC) is described. A first integrated circuit die is capable of a first operational mode. A second integrated circuit die is coupled to the first integrated circuit die. The first integrated circuit die has a rate of power consumption that is lower than the second integrated circuit die when the first integrated circuit die is in the first operational mode and the second integrated circuit die is in a second operational mode. The first integrated circuit die is configured for power management of the second integrated circuit die for placing the second integrated circuit die in a standby mode from the second operational mode and for returning the second integrated circuit die to the second operational mode from the standby mode.