US 11,734,059 B2
Hardware assisted fine-grained data movement
Muhammad Amber Hassaan, Austin, TX (US); Anirudh Mohan Kaushik, Austin, TX (US); Sooraj Puthoor, Austin, TX (US); Gokul Subramanian Ravi, Austin, TX (US); Bradford Beckmann, Bellevue, WA (US); and Ashwin Aji, Santa Clara, CA (US)
Assigned to Advanced Micro Devices, Inc., Santa Clara, CA (US)
Filed by ADVANCED MICRO DEVICES, INC., Santa Clara, CA (US)
Filed on Mar. 19, 2020, as Appl. No. 16/824,601.
Prior Publication US 2021/0294646 A1, Sep. 23, 2021
Int. Cl. G06F 9/46 (2006.01); G06F 9/48 (2006.01); G06F 9/52 (2006.01); G06F 16/901 (2019.01)
CPC G06F 9/4881 (2013.01) [G06F 9/52 (2013.01); G06F 16/9024 (2019.01); G06F 2209/486 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method implemented at a processor, comprising:
receiving, a task dependency graph representative of a plurality of tasks;
receiving task data requirements that correspond to each task of the plurality of tasks;
performing a task dependency assessment of the task dependency graph and the task data requirements; and
based on the task dependency assessment, scheduling for execution a first task of the plurality of tasks and a proxy object of a plurality of proxy objects specified by the task data requirements such that a memory transfer of the proxy object occurs while the first task is being executed.