Systems and methods for automated systolic array design from a high-level program are
disclosed. One implementa tion of a systolic array design supporting a convolutional neural …

(57) ABSTRACT A method for optimization of machine learning (ML) work loads on a
graphics processor unit (GPU). The method includes identifying a computation having a …

Abstract We present R-Stream ⋅ TF, a polyhedral optimization tool for neural network
computations. R-Stream ⋅ TF transforms computations performed in a neural network graph …

Data-parallel computation programs may be improved by, for example, determining the
functional properties user defined functions (UDFs), eliminating unnecessary data-shuffling …

In one general aspect, a computer system can include instruc tions stored on a non-
transitory computer-readable storage medium. The computer system can include a logging …

Systems and methods generate code from a source program where the generated code may
be compiled and executed on a Graphics Processing Unit (GPU). A parallel loop analysis …

The present invention relates to a method for compiling code for a multi-core processor,
comprising: detecting and opti mizing a loop, partitioning the loop into partitions execut able …

The present invention extends to methods, systems, and com puter program products for
automatically optimizing memory accesses by kernel functions executing on parallel …

a process virtual machine is generally understood to be a platform dependent engine that
executes platform independent code written in a high-level programming language for …

BACKGROUND 0003) A program that a computer is to be caused to execute is created, for
example, by using a high-level language, and is transformed into a computer-executable …