Abstract: Embodiments are generally directed to compression in machine learning and deep learning processing. An embodiment of an apparatus for compression of untyped data includes a graphical processing unit (GPU) including a data compression pipeline, the data compression pipeline including a data port coupled with one or more shader cores, wherein the data port is to allow transfer of untyped data without format conversion, and a 3D compression/decompression unit to provide for compression of untyped data to be stored to a memory subsystem and decompression of untyped data from the memory subsystem.

Abstract: Embodiments are generally directed to compression in machine learning and deep learning processing. An embodiment of an apparatus for compression of untyped data includes a graphical processing unit (GPU) including a data compression pipeline, the data compression pipeline including a data port coupled with one or more shader cores, wherein the data port is to allow transfer of untyped data without format conversion, and a 3D compression/decompression unit to provide for compression of untyped data to be stored to a memory subsystem and decompression of untyped data from the memory subsystem.

Abstract: Embodiments are generally directed to compression in machine learning and deep learning processing. An embodiment of an apparatus for compression of untyped data includes a graphical processing unit (GPU) including a data compression pipeline, the data compression pipeline including a data port coupled with one or more shader cores, wherein the data port is to allow transfer of untyped data without format conversion, and a 3D compression/decompression unit to provide for compression of untyped data to be stored to a memory subsystem and decompression of untyped data from the memory subsystem.

Abstract: Embodiments are generally directed to compression in machine learning and deep learning processing. An embodiment of an apparatus for compression of untyped data includes a graphical processing unit (GPU) including a data compression pipeline, the data compression pipeline including a data port coupled with one or more shader cores, wherein the data port is to allow transfer of untyped data without format conversion, and a 3D compression/decompression unit to provide for compression of untyped data to be stored to a memory subsystem and decompression of untyped data from the memory subsystem.

Abstract: The present disclosure is directed to systems and methods for decomposing systolic array circuitry to provide a plurality of N×N systolic sub-array circuits, apportioning a first tensor or array into a plurality of N×M first input arrays, and apportioning a second tensor or array into a plurality of M×N second input arrays. Systolic array control circuitry transfers corresponding ones of the first input arrays and second input arrays to a respective one of the plurality of N×N systolic sub-array circuits. As the elements included in the first input array and the elements included in the second input array are transferred to the systolic sub-array, the systolic sub-array performs one or more mathematical operations using the first and the second input arrays. The systems and methods beneficially improve the usage of the systolic array circuitry thereby advantageously reducing the number of clock cycles needed to perform a given number of calculations.

Abstract: The present disclosure is directed to systems and methods for decomposing systolic array circuitry to provide a plurality of N×N systolic sub-array circuits, apportioning a first tensor or array into a plurality of N×M first input arrays, and apportioning a second tensor or array into a plurality of M×N second input arrays. Systolic array control circuitry transfers corresponding ones of the first input arrays and second input arrays to a respective one of the plurality of N×N systolic sub-array circuits. As the elements included in the first input array and the elements included in the second input array are transferred to the systolic sub-array, the systolic sub-array performs one or more mathematical operations using the first and the second input arrays. The systems and methods beneficially improve the usage of the systolic array circuitry thereby advantageously reducing the number of clock cycles needed to perform a given number of calculations.

Abstract: Embodiments are generally directed to compression in machine learning and deep learning processing. An embodiment of an apparatus for compression of untyped data includes a graphical processing unit (GPU) including a data compression pipeline, the data compression pipeline including a data port coupled with one or more shader cores, wherein the data port is to allow transfer of untyped data without format conversion, and a 3D compression/decompression unit to provide for compression of untyped data to be stored to a memory subsystem and decompression of untyped data from the memory subsystem.

Abstract: Technologies for dynamic acceleration of general-purpose code include a computing device having a general-purpose processor core and one or more hardware accelerators. The computing device identifies an acceleration candidate in an application that is targeted to the processor core. The acceleration candidate may be a long-running computation of the application. The computing device translates the acceleration candidate into a translated executable targeted to the hardware accelerator. The computing device determines whether to offload execution of the acceleration candidate and, if so, executes the translated executable with the hardware accelerator. The computing device may translate the acceleration candidate into multiple translated executables, each targeted to a different hardware accelerator. The computing device may select among the translated executables in response to determining to offload execution.

Abstract: Systems, apparatuses, and methods for a hardware and software system to automatically decompose a program into multiple parallel threads are described. In some embodiments, the systems and apparatuses execute a method of original code decomposition and/or generated thread execution.

Abstract: Embodiments are generally directed to compression in machine learning and deep learning processing. An embodiment of an apparatus for compression of untyped data includes a graphical processing unit (GPU) including a data compression pipeline, the data compression pipeline including a data port coupled with one or more shader cores, wherein the data port is to allow transfer of untyped data without format conversion, and a 3D compression/decompression unit to provide for compression of untyped data to be stored to a memory subsystem and decompression of untyped data from the memory subsystem.

Abstract: Embodiments are generally directed to compression in machine learning and deep learning processing. An embodiment of an apparatus for compression of untyped data includes a graphical processing unit (GPU) including a data compression pipeline, the data compression pipeline including a data port coupled with one or more shader cores, wherein the data port is to allow transfer of untyped data without format conversion, and a 3D compression/decompression unit to provide for compression of untyped data to be stored to a memory subsystem and decompression of untyped data from the memory subsystem.

Abstract: An inter-architecture compatibility apparatus of an aspect includes a control flow transfer reception module to receive a first call procedure operation, intended for a first architecture library module, from a first architecture code module. The first call procedure operation involves a first plurality of input parameters. An application binary interface (ABI) change module is coupled with the control flow transfer reception module. The ABI change module makes ABI changes to convert the first call procedure operation involving the first plurality of input parameters to a corresponding second call procedure operation involving a second plurality of input parameters. The second call procedure operation is compatible with a second architecture library module. A control flow transfer output module is coupled with the ABI change module. The control flow transfer output module provides the second call procedure operation to the second architecture library module.

Abstract: Technologies for dynamic acceleration of general-purpose code include a computing device having a general-purpose processor core and one or more hardware accelerators. The computing device identifies an acceleration candidate in an application that is targeted to the processor core. The acceleration candidate may be a long-running computation of the application. The computing device translates the acceleration candidate into a translated executable targeted to the hardware accelerator. The computing device determines whether to offload execution of the acceleration candidate and, if so, executes the translated executable with the hardware accelerator. The computing device may translate the acceleration candidate into multiple translated executables, each targeted to a different hardware accelerator. The computing device may select among the translated executables in response to determining to offload execution.

Abstract: Systems, apparatuses, and methods for a hardware and software system to automatically decompose a program into multiple parallel threads are described. In some embodiments, the systems and apparatuses execute a method of original code decomposition and/or generated thread execution.

Abstract: A mechanism for tracking the control flow of instructions in an application and performing one or more optimizations of a processing device, based on the control flow of the instructions in the application, is disclosed. Control flow data is generated to indicate the control flow of blocks of instructions in the application. The control flow data may include annotations that indicate whether optimizations may be performed for different blocks of instructions. The control flow data may also be used to track the execution of the instructions to determine whether an instruction in a block of instructions is assigned to a thread, a process, and/or an execution core of a processor, and to determine whether errors have occurred during the execution of the instructions.

Abstract: Systems, apparatuses, and methods for a hardware and software system to automatically decompose a program into multiple parallel threads are described. In some embodiments, the systems and apparatuses execute a method of original code decomposition and/or generated thread execution.

Abstract: Systems, apparatuses, and methods for a hardware and software system to automatically decompose a program into multiple parallel threads are described. For example, a method according to one embodiment comprises: analyzing a single-threaded region of executing program code, the analysis including identifying dependencies within the single-threaded region; determining portions of the single-threaded region of executing program code which may be executed in parallel based on the analysis; assigning the portions to two or more parallel execution tracks; and executing the portions in parallel across the assigned execution tracks.

Abstract: Embodiments of computer-implemented methods, systems, computing devices, and computer-readable media (transitory and non-transitory) are described herein for analyzing execution of a plurality of executable instructions and, based on the analysis, providing an indication of a benefit to be obtained by vectorization of at least a subset of the plurality of executable instructions. In various embodiments, the analysis may include identification of the subset of the plurality of executable instructions suitable for conversion to one or more single-instruction multiple-data (“SIMD”) instructions.

Abstract: An inter-architecture compatibility apparatus of an aspect includes a control flow transfer reception module to receive a first call procedure operation, intended for a first architecture library module, from a first architecture code module. The first call procedure operation involves a first plurality of input parameters. An application binary interface (ABI) change module is coupled with the control flow transfer reception module. The ABI change module makes ABI changes to convert the first call procedure operation involving the first plurality of input parameters to a corresponding second call procedure operation involving a second plurality of input parameters. The second call procedure operation is compatible with a second architecture library module. A control flow transfer output module is coupled with the ABI change module. The control flow transfer output module provides the second call procedure operation to the second architecture library module.

Abstract: A mechanism for tracking the control flow of instructions in an application and performing one or more optimizations of a processing device, based on the control flow of the instructions in the application, is disclosed. Control flow data is generated to indicate the control flow of blocks of instructions in the application. The control flow data may include annotations that indicate whether optimizations may be performed for different blocks of instructions. The control flow data may also be used to track the execution of the instructions to determine whether an instruction in a block of instructions is assigned to a thread, a process, and/or an execution core of a processor, and to determine whether errors have occurred during the execution of the instructions.