IP Library › Granted Patent US 12,356,126
Granted Patent B2
US 12,356,126 · App. 18/520,339 · Granted Jul 8, 2025

Efficiently interconnecting computing nodes to enable use of high-radix network switches

Inventors: Hitesh Ballani (Cambridge, GB); Winston Allen Saunders (Seattle, WA); Christian L. Belady (Mercer Island, WA); Lisa Ru-Feng Hsu (Durham, NC); Paolo Costa (London, GB); Douglas M. Carmean (Seattle, WA)
Assignee: Microsoft Technology Licensing, LLC
H04Q11/0066H04Q11/0005H04Q11/0067H04Q11/0071H04Q2011/0026H04Q2011/0052H04Q2011/0079
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 12,356,126
App. No.
18/520,339
Granted
Jul 8, 2025
Kind
B2
Abstract

A system for efficiently interconnecting computing nodes can include a plurality of computing nodes and a plurality of network switches coupled in parallel to the plurality of computing nodes. The system can also include a plurality of node interfaces. Each computing node among the plurality of computing nodes can include at least one node interface for each network switch among the plurality of network switches. The plurality of node interfaces corresponding to a computing node can be configured to send data to another computing node via the plurality of network switches. The system can also include a plurality of switch interfaces. Each network switch among the plurality of network switches can include at least one switch interface for each computing node among the plurality of computing nodes. A switch interface corresponding to the computing node can be coupled to a node interface corresponding to the computing node.

Claims (43)

1. A method for efficiently processing a plurality of data streams while avoiding high-radix network devices, the method comprising:

sending, by a plurality of source node interfaces on a source node, a plurality of data streams, each data stream of the plurality of data streams being sent by a source node interface from the plurality of source node interfaces;

causing the plurality of data streams to be transmitted via a plurality of network switches coupled to the plurality of source node interfaces, each data stream being transmitted via a network switch from the plurality of network switches, and wherein each network switch from the plurality of network switches includes a first interface to receive data from the source node and a second interface to provide data to a destination node; and

receiving, at a plurality of destination node interfaces on the destination node, the plurality of data streams, each data stream of the plurality of data streams being received at a destination node interface from the plurality of destination node interfaces,

wherein the plurality of source node interfaces and the plurality of destination node interfaces send and receive the plurality of data streams without the source node performing serialization and without the destination node performing deserialization.

2. The method of claim 1 , wherein the plurality of network switches are coupled in parallel between the source node and the destination node.

3. The method of claim 1 , wherein each source node interface of the plurality of source node interfaces and each destination node interface of the plurality of destination node interfaces are configured to receive a serial data stream from a data source and send the serial data stream to another computing node via a network switch among the plurality of network switches.

4. The method of claim 1 , wherein:

the plurality of network switches are optically coupled to the source node and the destination node,

the plurality of source node interfaces and the plurality of destination node interfaces include a plurality of node optical transceivers, each source node interface and each destination node interface including at least one node optical transceiver, and

the first interface and the second interface on the plurality of network switches include a plurality of optical transceivers, each interface of the first interface and the second interface including at least one optical transceiver.

5. The method of claim 4 , wherein communications between the plurality of node optical transceivers and the plurality of optical transceivers occur via free-space optical communications.

6. The method of claim 1 , wherein the plurality of network switches are electrically coupled to the source node and the destination node.

7. The method of claim 1 , wherein the plurality of source node interfaces on the source node and the plurality of destination node interfaces on the destination node are optically coupled to the plurality of network switches via a plurality of optical cables.

8. A method for efficiently processing a plurality of data streams while avoiding high-radix network devices, the method comprising:

sending, by a first source node interface on a source node, a first data stream;

sending, by a second source node interface on the source node, a second data stream;

causing the first data stream to be transmitted via a first network switch coupled to the first source node interface, wherein the first network switch includes a first interface to receive data from the source node and a second interface to provide data to a destination node;

causing the second data stream to be transmitted via a second network switch coupled to the second source node interface, wherein the second network switch include a third interface to receive data from the source node and a fourth interface to provide data to the destination node;

receiving, at a first destination node interface on the destination node, the first data stream, wherein the first source node interface and the first destination node interface send and receive the first data stream without the source node performing serialization and without the destination node performing deserialization; and

receiving, at a second destination node interface on the destination node, the second data stream, wherein the second source node interface and the second destination node interface send and receive the second data stream without the source node performing serialization and without the destination node performing deserialization.

9. The method of claim 8 , wherein the first network switch and the second network switch are coupled in parallel between the source node and the destination node.

10. The method of claim 8 , wherein each node interface from the first source node interface, second source node interface, first destination node interface, and second destination node interface is configured to receive a serial data stream from a data source and send the serial data stream to another computing node via a network switch among the plurality of network switches.

11. The method of claim 8 , wherein:

the first network switch and the second network switch are optically coupled to the source node and the destination node,

the first source node interface, second source node interface, first destination node interface, and second destination node interface include a plurality of node optical transceivers, and

the first interface, second interface, third interface, and the fourth interface on the first network switch and the second network switch include a plurality of optical transceivers.

12. The method of claim 11 , wherein communications between the plurality of node optical transceivers and the plurality of optical transceivers occur via free-space optical communications.

13. The method of claim 8 , wherein the first network switch and the second network switch are electrically coupled to the source node and the destination node.

14. The method of claim 8 , wherein the first source node interface and the second source node interface on the source node and the first destination node interface and the second destination node interface on the destination node are optically coupled to the first network switch and the second network switch via a plurality of optical cables.

15. A system for efficiently processing a plurality of data streams while avoiding high-radix network devices, the system comprising:

a plurality of source node interfaces on a source node having a plurality of source node interfaces thereon, the plurality of source node interfaces being configured to send a plurality of data streams, each data stream of the plurality of data streams being sent by a source node interface from the plurality of source node interfaces;

a plurality of network switches coupled to the plurality of source node interfaces, each network switch from the plurality of network switches including a first interface to receive data from the source node and a second interface to provide data to a destination node, the plurality of network switches being configured to route the plurality of data streams from the source node to the destination node, wherein each data stream is transmitted via a network switch from the plurality of network switches; and

a plurality of destination node interfaces on the destination node, the plurality of destination node interfaces being configured to receive the plurality of data streams, each data stream from the plurality of data streams being received at a destination node interface from the plurality of destination node interfaces,

wherein the plurality of source node interfaces and the plurality of destination node interfaces send and receive the plurality of data streams without the source node performing serialization and without the destination node performing deserialization.

16. The system of claim 15 , wherein the plurality of network switches are coupled in parallel between the source node and the destination node.

17. The system of claim 15 , wherein each source node interface of the plurality of source node interfaces and each destination node interface of the plurality of destination node interfaces are configured to receive a serial data stream from a data source and send the serial data stream to another computing node via a network switch among the plurality of network switches.

18. The system of claim 15 , wherein:

the plurality of network switches are optically coupled to the source node and the destination node,

the plurality of source node interfaces and the plurality of destination node interfaces include a plurality of node optical transceivers, each source node interface and each destination node interface including at least one node optical transceiver, and

the first interface and the second interface on the plurality of network switches include a plurality of optical transceivers, each interface of the first interface and the second interface including at least one optical transceiver.

19. The system of claim 18 , wherein communications between the plurality of node optical transceivers and the plurality of optical transceivers occur via free-space optical communications.

20. The system of claim 18 , wherein the plurality of network switches are electrically coupled to the source node and the destination node.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Nov 27, 2023
From: BALLANI, HITESH; SAUNDERS, WINSTON ALLEN; BELADY, CHRISTIAN L.; HSU, LISA RU-FENG; COSTA, PAOLO; CARMEAN, DOUGLAS M.
To: MICROSOFT TECHNOLOGY LICENSING, LLC
Reel/Frame 065671/0948 →
Continuity (3)
Continuation 17243534 · Apr 28, 2021
Provisional Application 63109259 · Nov 3, 2020
Related Publication 20240098391A1 · Mar 21, 2024
References Cited (36)
US 4829511A · Georgiou · 1989 [cited by examiner]
US 5296950A · Lin et al. · 1994 [cited by applicant]
US 8659895B1 · Carlson et al. · 2014 [cited by applicant]
US 10491302B1 · Morris et al. · 2019 [cited by applicant]
US 11476934B1 · Saunders · 2022 [cited by applicant]
US 20040156640A1 · Dress et al. · 2004 [cited by applicant]
US 20080008471A1 · Dress · 2008 [cited by applicant]
US 20080008472A1 · Dress et al. · 2008 [cited by applicant]
US 20120155885A1 · Hannah · 2012 [cited by applicant]
US 20130182620A1 · Chaffee et al. · 2013 [cited by applicant]
US 20130223846A1 · Joseph · 2013 [cited by applicant]
US 20150078383A1 · Wisehart · 2015 [cited by applicant]
US 20150280827A1 · Adiletta et al. · 2015 [cited by applicant]
US 20160173199A1 · Gupta · 2016 [cited by applicant]
US 20170069214A1 · LeBlanc et al. · 2017 [cited by applicant]
US 20170237971A1 · Pitts · 2017 [cited by applicant]
US 20170257970A1 · Alleman et al. · 2017 [cited by applicant]
US 20180287818A1 · Goel · 2018 [cited by examiner]
US 20190235929A1 · Altstaetter et al. · 2019 [cited by applicant]
US 20190327780A1 · Ganguly · 2019 [cited by applicant]
US 20220368420A1 · Saunders et al. · 2022 [cited by applicant]
Chaintoutis, et al., “Free Space Intra-Datacenter Interconnects Based on 2D Optical Beam Steering Enabled by Photonic Integrated Circuits”, In Journal of Photonics, vol. 5, No. 3, Multidisciplinary Digital Publishing In… [cited by applicant]
Curran, et al., “Providing Line-of-Sight in a Free-Space-Optics Based Data Center Architecture”, In Proceedings IEEE International Conference on Communications (ICC), May 22, 2016, 7 Pages. [cited by applicant]
Fujiwara, et al_, “Augmenting Low-Latency HPC Network with Free-Space Optical Links”, In IEEE 21st International Symposium on High Performance Computer Architecture, Feb. 7, 2015, pp. 390-401. [cited by applicant]
Hamedazimi, et al., “Firefly: A Reconfigurable Wireless Data Center Fabric Using Free-Space Optics”, In Proceedings of the ACM conference on SIGCOMM, Aug. 17, 2014, pp. 319-330. [cited by applicant]
Hirabayashi, et al., “Optical Beam Direction Compensating System for Board-to-Board Free Space Optical Interconnection in High-Capacity ATM Switch”, In Journal of Lightwave technology, vol. 15, Issue No. 5, May 1997, pp… [cited by applicant]
Hu, et al., “Towards Tightly Coupled Datacenter with Free-Space Optical Links”, In Proceedings of the International Conference on Cloud and Big Data Computing, Sep. 17, 2017, pp. 33-39. [cited by applicant]
International Search Report and Written Opinion received for PCT Application No. PCT/US21/030770, Aug. 10, 2021, 14 pages. [cited by applicant]
Invitation pursuant to Rule 137(4) EPC and Article 94(3) Received for European Application No. 21736076.7, mailed on Jun. 6, 2024, 3 pages. [cited by applicant]
Lin, et al., “Free-Space Micromachined Optical Switches for Optical Networking”, In Journal of Selected Topics in Quantum Electronics, IEEE, vol. 5, Issue No. 1, Jan. 1999, pp. 4-9. [cited by applicant]
Non-Final Office Action Issued in U.S. Appl. No. 17/536,405, Mailed Date: Apr. 24, 2023, 21 Pages. [cited by applicant]
Non-Final Office Action mailed on Dec. 16, 2020, in U.S. Appl. No. 16/917,829, 19 pages. [cited by applicant]
Notice of Allowance mailed on Apr. 22, 2021, in U.S. Appl. No. 16/917,875, 6 pages. [cited by applicant]
Notice of Allowance mailed on Jun. 3, 2022, in U.S. Appl. No. 16/917,829, 15 pages. [cited by applicant]
Notice of Allowance mailed on Mar. 26, 2021, in U.S. Appl. No. 16/917,829, 14 pages. [cited by applicant]
Saljoghei, et al., “MCF-SMF Hybrid Low-Latency Circuit-Switched Optical Network for Disaggregated Data Centers”, In Journal of Lightwave Technology, vol. 37, Issue No. 16, Aug. 15, 2019, pp. 4017-4029. [cited by applicant]