Abstract: Techniques for data manipulation using a matrix multiplication engine using pipelining are disclosed. A first and a second matrix are obtained for matrix multiplication. A first matrix multiply-accumulate (MAC) unit is configured, where a first matrix element and a second matrix element are presented to the MAC unit on a first cycle. A second MAC unit is configured in pipelined fashion, where the first element of the first matrix and a second element of the second matrix are presented to the second MAC unit on a second cycle, and where a second element of the first matrix and the first element of the second matrix are presented to the first MAC unit on the second cycle. Additional MAC units are further configured within the processor in pipelined fashion. Multiply-accumulate operations are executed in pipelined fashion on each of n MAC units over additional k sets of m cycles.

Abstract: Techniques for a neural network output layer for machine learning are disclosed. A plurality of processing elements within a reconfigurable fabric is configured to implement a data flow graph, where the data flow graph implements a neural network. The data flow graph can include machine learning or deep learning. A layer is implemented, within the neural network, that maps a first vector of real values to a second vector of real values bounded by zero and one, where the second vector sums to a value of one using fixed-point calculations. The layer can include a final layer within the neural network. The layer that maps the first vector includes a Softmax function. Results of the neural network are classified based on a value of the second vector. The classifying can include part of a machine learning or a deep learning process.

Abstract: Techniques for data manipulation using processor cluster address generation are disclosed. One or more processor clusters capable of executing software-initiated work requests are accessed. A direct memory access (DMA) engine, coupled to the one or more processor clusters, is configured, wherein the DMA engine employs address generation across a plurality of tensor dimensions. A work request address field is parsed, where the address field contains unique address space descriptors for each of the plurality of dimensions, along with a common address space descriptor. DMA addresses are generated based on the unique address space descriptors and the common address space descriptor. Memory using two or more of the DMA addresses that were generated is accessed, where the two or more DMA addresses enable processing within the one or more processor clusters.

Abstract: Techniques are disclosed for data manipulation within a reconfigurable computing environment for data flow graph computation using exceptions. Processing elements are configured within a reconfigurable fabric to implement a data flow graph. The processing elements are loaded with process agents. Valid data is executed by a first process agent on a first processing element, where the first process agent corresponds to a starting node of the data flow graph. A second processing element detects that an error exception has occurred, where a second process agent is running on the second processing element. A done signal to a third process agent is withheld by the second process agent, where the third process agent is running on a third processing element. The second process agent raises an interrupt request, where the interrupt request is based on the detecting that an error exception has occurred.

Abstract: Techniques are disclosed for data manipulation. Data is obtained from a first switching element where the first switching element is controlled by a first circular buffer. Data is sent to a second switching element where the second switching element is controlled by a second circular buffer. Data is controlled by a third switching element that is controlled by a third circular buffer. The third switching element hierarchically controls the first switching element and the second switching element. Data is routed through a fourth switching element that is controlled by a fourth circular buffer. The circular buffers are statically scheduled. The obtaining data from a first switching element and the sending the data to a second switching element includes a direct memory access (DMA). The switching elements can operate as a master controller or as a slave device. The switching elements can comprise clusters within an asynchronous reconfigurable fabric.

Abstract: Techniques are disclosed for managing data within a reconfigurable computing environment. In a multiple processing element environment, such as a mesh network or other suitable topology, there is an inherent need to pass data between processing elements. Subtasks are divided among multiple processing elements. The output resulting from the subtasks is then merged by a downstream processing element. In such cases, a join operation can be used to combine data from multiple upstream processing elements. A control agent executes on each processing element. A memory buffer is disposed between upstream processing elements and the downstream processing element. The downstream processing element is configured to automatically perform an operation based on the availability of valid data from the upstream processing elements.

Abstract: Techniques are disclosed for power conservation. A plurality of processing elements and a plurality of instructions are configured. The plurality of processing elements is controlled by instructions contained in a plurality of circular buffers. The plurality of processing elements can comprise a data flow processor. A first processing element, from the plurality of interconnected processing elements, is set into a sleep state by a first instruction from the plurality of instructions. The first processing element is woken from the sleep state as a result of valid data being presented to the first processing element. A subsection of the plurality of interconnected processing elements is also set into a sleep state based on the first processing element being set into a sleep state.

Abstract: A plurality of software programmable processors is disclosed. The software programmable processors are controlled by rotating circular buffers. A first processor and a second processor within the plurality of software programmable processors are individually programmable. The first processor within the plurality of software programmable processors is coupled to neighbor processors within the plurality of software programmable processors. The first processor sends and receives data from the neighbor processors. The first processor and the second processor are configured to operate on a common instruction cycle. An output of the first processor from a first instruction cycle is an input to the second processor on a subsequent instruction cycle.

Abstract: A combination of memory units and dataflow processing units is disclosed for computation. A first memory unit is interposed between a first dataflow processing unit and a second dataflow processing unit. Operations for a dataflow graph are allocated across the first dataflow processing unit and the second dataflow processing unit. The first memory unit passes data between the first dataflow processing unit and the second dataflow processing unit to execute the dataflow graph. The first memory unit is a high bandwidth, shared memory device including a hybrid memory cube. The first dataflow processing unit and second dataflow processing unit include a plurality of circular buffers containing instructions for controlling data transfer between the first dataflow processing unit and second dataflow processing unit. Additional dataflow processing units and additional memory units are included for additional functionality and efficiency.

Abstract: Disclosed embodiments provide an interface circuit for the transfer of data from a synchronous circuit to an asynchronous circuit. Data from the synchronous circuit is received into a memory in the interface circuit. The data in the memory is then sent to the asynchronous circuit based on an instruction in a circular buffer that is part of the interface circuit. Processing elements within the interface circuit execute instructions contained within the circular buffer. The circular buffer rotates to provide new instructions to the processing elements. Flow control paces the data from the synchronous circuit to the asynchronous circuit.

Abstract: Disclosed techniques utilize a satisfiability solver for allocation and/or configuration of resources in a reconfigurable fabric of processing elements. A dataflow graph is an input provided to a toolchain that includes a satisfiability solver. The satisfiability solver operates on subsets of interconnected nodes within a dataflow graph to derive a solution. The solution is trimmed by removing artifacts and unnecessary parts. The solutions of subsets are then used as an input to additional subsets of nodes within the dataflow graph in an iterative process to derive a complete solution. The satisfiability solver technique uses adaptive windowing in both the time dimension and the spatial dimensions of the dataflow graph. Processing elements and routing elements within the reconfigurable fabric are configured based on the complete solution. Data computation is performed based on the dataflow graph using the processing elements and the routing resources.

Abstract: Circular buffers containing instructions that enable the execution of operations on logical elements are described where data in the circular buffers is swapped to storage. The instructions comprise a branchless instruction set. Data stored in circular buffers is paged in and out to a second level memory. State information for each logical element is also saved and restored using paging memory. Instructions are provided to logical elements, such as processing elements, via circular buffers. The instructions enable a group of processing elements to perform operations implementing a desired functionality. That functionality is changed by updating the circular buffers with new instructions that are transferred from paging memory. The previous instructions can be saved off in paging memory before the new instructions are copied over to the circular buffers. This enables the hardware to be rapidly reconfigured amongst multiple functions.

Abstract: Techniques are disclosed for designing a reconfigurable fabric. The reconfigurable fabric is designed using logical elements, configurable connections between and among the logical elements, and rotating circular buffers. The circular buffers contain configuration instructions. The configuration instructions control connections between and among logical elements. The logical elements change operation based on the instructions that rotate through the circular buffers. Clusters of logical elements are interconnected by a switching fabric. Each cluster contains processing elements, storage elements, and switching elements. A circular buffer within a cluster contains multiple switching instructions to control the flow of data throughout the switching fabric. The circular buffer provides a pipelined execution of switching instructions for the implementation of multiple functions.

Abstract: An interface circuit is disclosed for the transfer of data from a synchronous circuit, with multiple source elements, to an asynchronous circuit. Data from the synchronous circuit is received into a memory in the interface circuit. The data in the memory is then sent to the asynchronous circuit based on an instruction in a circular buffer that is part of the interface circuit. Processing elements within the interface circuit execute instructions contained within the circular buffer. The circular buffer rotates to provide new instructions to the processing elements. Flow control paces the data from the synchronous circuit to the asynchronous circuit.

Abstract: Methods and systems for timing analysis and optimization of asynchronous circuit designs are disclosed. Registration stages are placed between combinational logic circuits. For timing purposes, the registration stages are modified to have a duplicate set of pins. New paths are formed in the circuit for the purposes of timing analysis. The paths are analyzable by timing tools. Once the timing analysis is complete, the paths are reverted to original paths, and new devices are selected for the circuit design based on results of the timing analysis. An updated design is sent for manufacture, based on the timing analysis and optimization of the asynchronous circuit.

Abstract: An apparatus for mathematical manipulation is described allowing the selective combination of shifters to shift binary numbers of various widths. Selective combination allows on-the-fly adjustment of shifters from independent to coordinated shifting operations. Selective combination allows adjustable hardware-based shifting while saving space and resources. Multiple eight-bit shifters can be configured for a variety of operand widths, such as a 32-bit width, a 24-bit width, a 16-bit width, or an eight-bit width. Multiplexers route the appropriate input data to the appropriate shifters. Bidirectional shifting is configured through a selector tree, including both shift left and shift right operations. Opcodes configure the shifters for the desired type of shift and a shifted result is generated.

Abstract: Clusters of logical elements are interconnected by a switching fabric. Each cluster contains processing elements, storage elements, and switching elements. A circular buffer within a cluster contains multiple switching instructions to control the flow of data throughout the switching fabric. The circular buffer provides a pipelined execution of switching instructions for the implementation of multiple functions. Each cluster contains multiple processing elements, and each cluster further comprises an additional circular buffer for each processing element. Logical operations are controlled by the circular buffers.

Abstract: Techniques are disclosed for power conservation. A plurality of processing elements and a plurality of instructions are configured. The plurality of processing elements is controlled by instructions contained in a plurality of circular buffers. The plurality of processing elements can comprise a dataflow processor. A first processing element, from the plurality of interconnected processing elements, is set into a sleep state by a first instruction from the plurality of instructions. The first processing element is woken from the sleep state as a result of valid data being presented to the first processing element. A subsection of the plurality of interconnected processing elements is also set into a sleep state based on the first processing element being set into a sleep state.

Abstract: Circular buffers containing instructions that enable the execution of operations on logical elements are described where data in the circular buffers is swapped to storage. Data stored in circular buffers is paged in and out to a second level memory. State information for each logical element is also saved and restored using paging memory. Logical elements such as processing elements are provided instructions via circular buffers. The instructions enable a group of processing elements to perform operations implementing a desired functionality. That functionality is changed by updating the circular buffers with new instructions that are transferred from paging memory. The previous instructions can be saved off in paging memory before the new instructions are copied over to the circular buffers. This enables the hardware to be rapidly reconfigured amongst multiple functions.

Abstract: Systems and methods are disclosed for computing resource allocation based on flow graph translation. First, a high-level description of logic circuitry is obtained and translated to generate a flow graph representing sequential operations. Using the flow graph, similar processing elements in an array are interchangeably allocated to perform computational, communication, and storage tasks as needed. The sequential operations are executed using the array of interchangeable processing elements. Data is provided from the storage elements through the communication elements to the computational elements. Computational results are stored in the storage elements. Outputs from some of the computational elements provide inputs to other computational elements. Execution of the instructions can be controlled with time stepping. The processors are reallocated as needed, based on changes to the flow graph.