DOE Patents title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: GPU remote communication with triggered operations

Abstract

Methods, devices, and systems for transmitting data over a computer communications network are disclosed. A queue of communications commands can be pre-generated using a central processing unit (CPU) and stored in a device memory of a network interface controller (NIC). Thereafter, if a graphics processing unit (GPU) has data to communicate to a remote GPU, it can store the data in a send buffer, where the location in the buffer is pointed to by a pre-generated command. The GPU can then signal to the interface device that the data is ready, triggering execution of the pre-generated command to send the data.

Inventors:
;
Issue Date:
Research Org.:
Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1805585
Patent Number(s):
10936533
Application Number:
15/297,079
Assignee:
Advanced Micro Devices, Inc. (Sunnyvale, CA)
Patent Classifications (CPCs):
G - PHYSICS G06 - COMPUTING G06F - ELECTRIC DIGITAL DATA PROCESSING
H - ELECTRICITY H04 - ELECTRIC COMMUNICATION TECHNIQUE H04L - TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
DOE Contract Number:  
AC02-05CH11231
Resource Type:
Patent
Resource Relation:
Patent File Date: 10/18/2016
Country of Publication:
United States
Language:
English

Citation Formats

LeBeane, Michael W., and Reinhardt, Steven K. GPU remote communication with triggered operations. United States: N. p., 2021. Web.
LeBeane, Michael W., & Reinhardt, Steven K. GPU remote communication with triggered operations. United States.
LeBeane, Michael W., and Reinhardt, Steven K. Tue . "GPU remote communication with triggered operations". United States. https://www.osti.gov/servlets/purl/1805585.
@article{osti_1805585,
title = {GPU remote communication with triggered operations},
author = {LeBeane, Michael W. and Reinhardt, Steven K.},
abstractNote = {Methods, devices, and systems for transmitting data over a computer communications network are disclosed. A queue of communications commands can be pre-generated using a central processing unit (CPU) and stored in a device memory of a network interface controller (NIC). Thereafter, if a graphics processing unit (GPU) has data to communicate to a remote GPU, it can store the data in a send buffer, where the location in the buffer is pointed to by a pre-generated command. The GPU can then signal to the interface device that the data is ready, triggering execution of the pre-generated command to send the data.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Tue Mar 02 00:00:00 EST 2021},
month = {Tue Mar 02 00:00:00 EST 2021}
}

Works referenced in this record:

Enforcing transaction order in peer-to-peer interactions
patent-application, December 2016


GPU and Encoding Apparatus for Virtual Machine Environments
patent-application, September 2014


Low-Latency Processing in a Network Node
patent-application, October 2015


Remote Task Queuing by Networked Computing Devices
patent-application, November 2014


Integrated GPU, NIC and Compression Hardware for Hosted Graphics
patent-application, January 2010


Technologies for Offloading Network Packet Processing to a GPU
patent-application, March 2017


Apparatus and Method for Memory-Hierarchy Aware Producer-Consumer Instruction
patent-application, July 2014


Collaborative Hardware Interaction by Multiple Entities Using a Shared Queue
patent-application, April 2016


Rendering Content on Computing Systems
patent-application, July 2015


Graphics Processing Hardware for Using Compute Shaders as Front End for Vertex Shaders
patent-application, December 2014


Communication between integrated graphics processing units
patent, August 2018


Optimizing Resolve Performance with Tiling Graphics Architectures
patent-application, March 2013


GPU Operation
patent-application, November 2016


Adaptive Content Switching
patent-application, August 2017


Dynamic pinning remote direct memory access
patent, March 2012


Graphics Processing Unit (GPU) as a Programmable Packet Transfer Mechanism
patent-application, January 2018


Heterogeneous Input/Output (I/O) Using Remote Direct Memory Access (RDMA) and Active Message
patent-application, June 2016


Dynamic Foveation Adjustment
patent-application, February 2018