US 12,730,617 B2
Dataflow graph programming environment for a heterogenous processing system
Shail Aditya Gupta, San Jose, CA (US); Samuel R. Bayliss, Mountain View, CA (US); Vinod Kathail, Palo Alto, CA (US); Ralph D. Wittig, Menlo Park, CA (US); Philip B. James-Roxby, Longmont, CO (US); and Akella Sastry, San Jose, CA (US)
Assigned to XILINX, INC., San Jose, CA (US)
Filed by XILINX, INC., San Jose, CA (US)
Filed on Nov. 2, 2021, as Appl. No. 17/517,471.
Application 17/517,471 is a continuation of application No. 16/420,831, filed on May 23, 2019, granted, now 11,204,745.
Prior Publication US 2022/0058005 A1, Feb. 24, 2022
Int. Cl. G06F 9/44 (2018.01); G06F 8/41 (2018.01); G06F 9/54 (2006.01); G06F 15/78 (2006.01); G06F 16/901 (2019.01)
CPC G06F 8/433 (2013.01) [G06F 8/47 (2013.01); G06F 9/544 (2013.01); G06F 15/7807 (2013.01); G06F 16/9024 (2019.01)] 19 Claims
OG exemplary drawing
 
1. A method, comprising:
providing a heterogeneous programming environment for defining a dataflow graph as object oriented source code;
receiving graph source code generated in the heterogeneous programming environment, the graph source code defining a plurality of kernels and a plurality of communication links, wherein each of the plurality of communication links couple a respective pair of the plurality of kernels to form the dataflow graph;
compiling the graph source code to implement the dataflow graph in a heterogeneous processing system using compiled source code, wherein compiling the graph source code comprises:
assigning the plurality of kernels to the heterogeneous processing system based on constraints defined in the graph source code for the plurality of kernels, wherein the constraints are user-defined constraints that instruct the compiler to place a first kernel of the kernels in a first type of hardware element in the heterogeneous processing system and place a second kernel of the kernels in a second, different type of hardware element in the heterogeneous processing system,
assigning a communication type to the plurality of communication links defined in the graph source code, and
selecting synchronization techniques for transferring data between the plurality of kernels using the plurality of communication links; and
configuring the heterogeneous processing system to execute the dataflow graph using the compiled source code.