IP Library Granted Patent US 12,739,313
Granted Patent B2
US 12,739,313 · App. 18/090,686 · Granted Sep 15, 2026

Federated distribution of computation and operations using networked processing units

Inventors: Kshitij Arun Doshi (Tempe, AZ); Vesh Raj Sharma Banjade (Portland, OR); Satish Chandra Jha (Portland, OR); Francesc Guim Bernat (Barcelona, ES); S M Iftekharul Alam (Hillsboro, OR)
Assignee: Intel Corporation
H04L67/63G06F9/3005G06F9/3885G06F9/4881G06F9/5038G06F9/505G06F9/5072G06F9/5077G06F9/5083G06F9/54G06F11/0793G06F12/0851G06F12/0873H04L63/0876H04L63/12H04L67/10H04L67/1091H04L67/12H04L41/5003
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,739,313
App. No.
18/090,686
Granted
Sep 15, 2026
Kind
B2
Abstract

Various approaches for deploying and controlling distributed compute operations with the use of infrastructure processing units (IPUs) and similar network-addressable processing units are disclosed. A device for orchestrating functions in a network compute mesh is configured to receive, at a network-addressable processing unit of a network-addressable processing unit mesh from a requestor device, a computation request to execute a workflow with a set of objectives; query at least one other network-addressable processing units of the network-addressable processing unit mesh using the set of objectives, to determine aspects of available resources and data in the network-addressable processing unit mesh to apply to the workflow; transmit a list of recommended resources available to execute the workflow to the requestor device, the list of recommended resources being ranked based on at least one dimension of the resources; obtain a compute chain from the requestor device, the compute chain describing resource control transitions and data flow provided from the recommended resources and data in the network-addressable processing unit mesh; and schedule the execution of the workflow at one or more network-addressable processing units in the network-addressable processing unit mesh in accordance with the compute chain.

Claims (54)

1 . A device for orchestrating functions in a network compute mesh, comprising:

a memory device configured to store instructions; and

a processor subsystem, which when configured by the instructions, is operable to:

receive, at a network-addressable processing unit of a network-addressable processing unit mesh from a requestor device, a computation request to execute a workflow with a set of objectives;

query at least one other network-addressable processing units of the network- addressable processing unit mesh using the set of objectives, to determine aspects of available resources and data in the network-addressable processing unit mesh to apply to the workflow;

transmit a list of recommended resources available to execute the workflow to the requestor device, the list of recommended resources being ranked based on at least one dimension of the available resources;

obtain a compute chain from the requestor device, the compute chain describing resource control transitions and data flow provided from resources in the list of recommended resources and data in the network-addressable processing unit mesh; and

schedule the execution of the workflow at one or more network-addressable processing units in the network-addressable processing unit mesh in accordance with the compute chain.

2 . The device of claim 1 , wherein the set of objectives are expressed as service level objectives.

3 . The device of claim 1 , wherein the set of objectives are expressed as a multi-objective function.

4 . The device of claim 1 , wherein the set of objectives are default objectives.

5 . The device of claim 1 , wherein the aspects of available resources is provided by a second network-addressable processing unit of the network-addressable processing unit mesh in a resource map.

6 . The device of claim 1 , wherein the aspects of available resources include a percentage of compute, a number of cycles of compute, an amount of memory, an amount of storage, or network resources of a second network-addressable processing unit or a host managed by the network-addressable processing unit.

7 . The device of claim 1 , wherein the aspects of available data is provided by a second network-addressable processing unit of the network-addressable processing unit mesh in a data map.

8 . The device of claim 1 , wherein the aspects of available data include a location, a version, a type, or an amount or data.

9 . The device of claim 1 , wherein the processor subsystem is to:

receive a revised set of objectives from the requestor device;

query at least one other network-addressable processing units of the network-addressable processing unit mesh using the revised set of objectives, to determine aspects of available resources and data in the network-addressable processing unit mesh to apply to the workflow; and

transmit revised recommended resources available to execute the workflow to the requestor device, the revised recommended resources including a revised ranked list of resources based on at least one dimension of the available resources.

10 . The device of claim 1 , wherein the list of recommended resources includes a top N of resources based on the at least one dimension of the available resources.

11 . The device of claim 1 , wherein to schedule the execution of the workflow across the network-addressable processing unit mesh in accordance with the compute chain, the processor subsystem is to:

transmit the compute chain to each network-addressable processing unit in the network-addressable processing unit mesh that is assigned to a resource used in the compute chain, wherein the respective network-addressable processing units associated with the respective resources used in the compute chain cooperatively coordinate resource scheduling and data movements to execute the compute chain.

12 . The device of claim 1 , wherein intermediate results of the execution of the compute chain are stored in a logging database.

13 . The device of claim 1 , wherein the execution of the compute chain produces a result, which is stored in a logging database.

14 . A method for orchestrating functions in a network compute mesh, comprising:

receiving, at a network-addressable processing unit of a network-addressable processing unit mesh from a requestor device, a computation request to execute a workflow with a set of objectives;

querying at least one other network-addressable processing units of the network-addressable processing unit mesh using the set of objectives, to determine aspects of available resources and data in the network-addressable processing unit mesh to apply to the workflow;

transmitting a list of recommended resources available to execute the workflow to the requestor device, the list of recommended resources being ranked based on at least one dimension of the available resources;

obtaining a compute chain from the requestor device, the compute chain describing resource control transitions and data flow provided from resources in the list of recommended resources and data in the network-addressable processing unit mesh; and

scheduling the execution of the workflow at one or more network-addressable processing units in the network-addressable processing unit mesh in accordance with the compute chain.

15 . The method of claim 14 , wherein the set of objectives are expressed in a service level agreement.

16 . The method of claim 14 , comprising:

receiving a revised set of objectives from the requestor device;

querying at least one other network-addressable processing units of the network-addressable processing unit mesh using the revised set of objectives, to determine aspects of available resources and data in the network-addressable processing unit mesh to apply to the workflow; and

transmitting revised recommended resources available to execute the workflow to the requestor device, the revised recommended resources including a revised ranked list of resources based on at least one dimension of the available resources.

17 . The method of claim 14 , wherein the list of recommended resources includes a top N of resources based on the at least one dimension of the available resources.

18 . The method of claim 14 , wherein scheduling the execution of the workflow across the network-addressable processing unit mesh in accordance with the compute chain comprises:

transmitting the compute chain to each network-addressable processing unit in the network-addressable processing unit mesh that is assigned to a resource used in the compute chain, wherein the respective network-addressable processing units associated with the respective resources used in the compute chain cooperatively coordinate resource scheduling and data movements to execute the compute chain.

19 . The method of claim 14 , wherein intermediate results of the execution of the compute chain are stored in a logging database.

20 . The method of claim 14 , wherein the execution of the compute chain produces a result, which is stored in a logging database.

21 . At least one non-transitory machine-readable medium including instructions for orchestrating functions in a network compute mesh, which when executed by a machine, cause the machine to:

receive, at a network-addressable processing unit of a network-addressable processing unit mesh from a requestor device, a computation request to execute a workflow with a set of objectives;

query at least one other network-addressable processing units of the network-addressable processing unit mesh using the set of objectives, to determine aspects of available resources and data in the network-addressable processing unit mesh to apply to the workflow;

transmit a list of recommended resources available to execute the workflow to the requestor device, the list of recommended resources being ranked based on at least one dimension of the available resources;

obtain a compute chain from the requestor device, the compute chain describing resource control transitions and data flow provided from resources in the list of recommended resources and data in the network-addressable processing unit mesh; and

schedule the execution of the workflow at one or more network-addressable processing units in the network-addressable processing unit mesh in accordance with the compute chain.

22 . The at least one non-transitory machine-readable medium of claim 21 , comprising instructions to:

receive a revised set of objectives from the requestor device;

query at least one other network-addressable processing units of the network-addressable processing unit mesh using the revised set of objectives, to determine aspects of available resources and data in the network-addressable processing unit mesh to apply to the workflow; and

transmit revised recommended resources available to execute the workflow to the requestor device, the revised recommended resources including a revised ranked list of resources based on at least one dimension of the available resources.

23 . The at least one non-transitory machine-readable medium of claim 21 , wherein the list of recommended resources includes a top N of resources based on the at least one dimension of the available resources.

24 . The at least one non-transitory machine-readable medium of claim 21 , wherein the instructions to schedule the execution of the workflow across the network-addressable processing unit mesh in accordance with the compute chain comprise instructions to:

transmit the compute chain to each network-addressable processing unit in the network-addressable processing unit mesh that is assigned to a resource used in the compute chain, wherein the respective network-addressable processing units associated with the respective resources used in the compute chain cooperatively coordinate resource scheduling and data movements to execute the compute chain.

25 . The at least one non-transitory machine-readable medium of claim 21 , wherein intermediate results of the execution of the compute chain are stored in a logging database.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 14, 2023
From: DOSHI, KSHITIJ ARUN; SHARMA BANJADE, VESH RAJ; JHA, SATISH CHANDRA; GUIM BERNAT, FRANCESC; ALAM, S M IFTEKHARUL
To: INTEL CORPORATION
Reel/Frame 062697/0080 →
Continuity (2)
Provisional Application 63425857 · Nov 16, 2022
Related Publication 20230136048A1 · May 4, 2023
References Cited (104)
US 7065764B1 · Prael · 2006 [cited by examiner]
US 7716336B2 · Coppinger · 2010 [cited by examiner]
US 7861246B2 · Lu · 2010 [cited by examiner]
US 9363087B2 · Hawblitzel et al. · 2016 [cited by applicant]
US 9811380B1 · Martin et al. · 2017 [cited by applicant]
US 9906493B1 · Rodgers et al. · 2018 [cited by applicant]
US 10044696B2 · Walker et al. · 2018 [cited by applicant]
US 10389800B2 · Blainey et al. · 2019 [cited by applicant]
US 11210218B1 · Tavallaei et al. · 2021 [cited by applicant]
US 11288096B2 · Ugale et al. · 2022 [cited by applicant]
US 11675630B2 · Behar et al. · 2023 [cited by applicant]
US 11687376B2 · Cheng et al. · 2023 [cited by applicant]
US 11924060B2 · Smith et al. · 2024 [cited by applicant]
US 20090279701A1 · Moisand et al. · 2009 [cited by applicant]
US 20110213508A1 · Mandagere et al. · 2011 [cited by applicant]
US 20110239220A1 · Gibson et al. · 2011 [cited by applicant]
US 20130261826A1 · Mandagere et al. · 2013 [cited by applicant]
US 20140143781A1 · Yao · 2014 [cited by applicant]
US 20150081893A1 · Chadwell et al. · 2015 [cited by applicant]
US 20150089507A1 · Nagao · 2015 [cited by examiner]
US 20150200854A1 · Buchnik et al. · 2015 [cited by applicant]
US 20150229582A1 · Zhu et al. · 2015 [cited by applicant]
US 20150268985A1 · Jokinen et al. · 2015 [cited by applicant]
US 20150350089A1 · Wang et al. · 2015 [cited by applicant]
US 20160004571A1 · Smith · 2016 [cited by applicant]
US 20160350157A1 · Necas · 2016 [cited by applicant]
US 20160380819A1 · Burger · 2016 [cited by applicant]
US 20170286122A1 · Wu et al. · 2017 [cited by applicant]
US 20180285766A1 · Shen et al. · 2018 [cited by applicant]
US 20180373863A1 · Domke · 2018 [cited by applicant]
US 20190034716A1 · Kamarol et al. · 2019 [cited by applicant]
US 20190043201A1 · Strong et al. · 2019 [cited by applicant]
US 20190116128A1 · Guo et al. · 2019 [cited by applicant]
US 20190319807A1 · Fairfax et al. · 2019 [cited by applicant]
US 20190327328A1 · Smith et al. · 2019 [cited by applicant]
US 20190370076A1 · Behar et al. · 2019 [cited by applicant]
US 20190370084A1 · Behar et al. · 2019 [cited by applicant]
US 20200065098A1 · Parandeh Afshar et al. · 2020 [cited by applicant]
US 20200065290A1 · Natu · 2020 [cited by applicant]
US 20200136906A1 · Guim Bernat et al. · 2020 [cited by applicant]
US 20200250003A1 · Yang et al. · 2020 [cited by applicant]
US 20200285523A1 · Guim Bernat · 2020 [cited by examiner]
US 20200372013A1 · Lee et al. · 2020 [cited by applicant]
US 20200394045A1 · Iwata · 2020 [cited by examiner]
US 20210011765A1 · Doshi et al. · 2021 [cited by applicant]
US 20210097108A1 · Goyal et al. · 2021 [cited by applicant]
US 20210109785A1 · Prabhakaran et al. · 2021 [cited by applicant]
US 20210200592A1 · Guim et al. · 2021 [cited by applicant]
US 20220046670A1 · Lin · 2022 [cited by applicant]
US 20220164242A1 · Sui et al. · 2022 [cited by applicant]
US 20220206849A1 · Guim Bernat et al. · 2022 [cited by applicant]
US 20220286481A1 · Shaw et al. · 2022 [cited by applicant]
US 20220400085A1 · Ananthanarayanan et al. · 2022 [cited by applicant]
US 20230004433A1 · Kan et al. · 2023 [cited by applicant]
US 20230132992A1 · Bernat et al. · 2023 [cited by applicant]
US 20230133020A1 · Sakalley et al. · 2023 [cited by applicant]
US 20230134683A1 · Carranza et al. · 2023 [cited by applicant]
US 20230135645A1 · Guim Bernat et al. · 2023 [cited by applicant]
US 20230135938A1 · Carranza et al. · 2023 [cited by applicant]
US 20230136612A1 · Doshi et al. · 2023 [cited by applicant]
US 20230136615A1 · Guim Bernat et al. · 2023 [cited by applicant]
US 20230137879A1 · Alam et al. · 2023 [cited by applicant]
US 20230140252A1 · Carranza et al. · 2023 [cited by applicant]
US 20230199440A1 · Lee et al. · 2023 [cited by applicant]
US 20230342478A1 · Ranganathan et al. · 2023 [cited by applicant]
US 20230401099A1 · Bruno et al. · 2023 [cited by applicant]
US 20230409233A1 · Sheeley et al. · 2023 [cited by applicant]
US 20240020173A1 · Chatterjee · 2024 [cited by applicant]
US 20240031221A1 · Rajagopalan et al. · 2024 [cited by applicant]
US 20240112076A1 · Jain · 2024 [cited by examiner]
US 20240127392A1 · Hughes et al. · 2024 [cited by applicant]
M. Z. Nayyer et al., “LBRO: Load Balancing for Resource Optimization in Edge Computing,” in IEEE Access, vol. 10, pp. 97439-97449, 2022, doi: 10.1109/ACCESS.2022.3205741. (Year: 2022). [cited by examiner]
Hu, Shuihai, “Deadlocks in Datacenter Networks: Why Do They Form, and How to Avoid Them”, HotNets-XV, (Nov. 2016), 7 pgs. [cited by applicant]
Kumar, Gautam, “Swift: Delay is Simple and Effective for Congestion Control in the Datacenter”, SIGCOMM '20, (Aug. 2020), 15 pgs. [cited by applicant]
Mittal, Radhika, “Timely: RTT-Based Congestion Control for the Data Center”, SIGCOMM, (Aug. 2015), 14 pgs. [cited by applicant]
“Elastics.cloud First to DemonstrateCXLtm Enabled Symmetric Multi HostMemory Pooling and Expansion”, [Online]. Retrieved from the Internet https www.prnewswire.com news-releases elasticscloud-first-to-demonstrate-cxl-en… [cited by applicant]
“The GigaIO Fabrex Memory Fabric Memory Pooling Overview”, [Online]. Retrieved from the Internet https gigaio.com project memory-pooling-overview , Feb. 2020, 8 pages. [cited by applicant]
“Marvell Announces Innovative CXL Development Platform for Multi-Host Memory Pooling”, [Online]. Retrieved from the Internet https www.prnewswire.com news-releases marvell-announces-innovative-cxl-development-platform-f… [cited by applicant]
“Intel Threading Building Blocks Tutorial v1.21”, Intel Corporation, Oct. 27, 2011, 90 pages. [cited by applicant]
“U.S. Appl. No. 18/090,749, Non Final Office Action mailed Jan. 8, 2026”, 11 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,720, Non Final Office Action mailed Jan. 28, 2026”, 13 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,786, Non Final Office Action mailed Jan. 28, 2026”, 21 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,701, Non Final Office Action mailed Feb. 9, 2026”, 9 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,764, Non Final Office Action mailed Feb. 11, 2026”, 15 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,813, Non Final Office Action mailed Feb. 13, 2026”, 17 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,653, Non Final Office Action mailed Feb. 20, 2026”, 25 pgs. [cited by applicant]
Ananthanarayanan, Ganesh, “Reining in the Outliers in Map Reduce Clusters using Mantri”, 9th USENIX Symposium on Operating Systems Design and Implementation OSDI 10., 2010, 14 pages. [cited by applicant]
Moore, Andrew, “IPU-Based Cloud Infrastructure: The Fulcrum for Digital Business” Intel Corporation, 2021, 6 pages. [cited by applicant]
“U.S. Appl. No. 18/090,749, Response filed Apr. 7, 2026 to Non Final Office Action mailed Jan. 8, 2026”, 13 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,862, Notice of Allowance mailed Apr. 13, 2026”, 11 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,720, Response filed Apr. 16, 2026 to Non Final Office Action mailed Jan. 28, 2026”, 12 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,842, Non Final Office Action mailed Apr. 20, 2026”, 17 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,786, Response filed Apr. 23, 2026 to Non Final Office Action mailed Jan. 28, 2026”, 11 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,862, Corrected Notice of Allowability mailed Apr. 27, 2026”, 2 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,749, Final Office Action mailed May 4, 2026”, 11 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,701, Response filed May 7, 2026 to Non Final Office Action mailed Feb. 9, 2026”, 12 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,764, Response filed May 7, 2026 to Non Final Office Action mailed Feb. 11, 2026”, 11 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,813, Response filed May 13, 2026 to Non Final Office Action mailed Feb. 13, 2026”, 9 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,653, Response filed May 20, 2026 to Non Final Office Action mailed Feb. 20, 2026”, 14 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,786, Notice of Allowance mailed May 22, 2026”, 9 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,701, Final Office Action mailed May 29, 2026”, 12 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,764, Final Office Action mailed Jun. 3, 2026”, 8 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,813, Notice of Allowance mailed Jun. 9, 2026”, 9 pgs. [cited by applicant]
“U.S. Appl. No. 18/090,720, Final Office Action mailed Jul. 1, 2026”, 14 pgs. [cited by applicant]