UnfoldIPThemes

Company patents

Innovium, Inc.

Innovium, Inc. exhibits a patent strategy heavily concentrated in communications, with Routing, Switching & QoS representing a dominant 78.7% of its portfolio, yet this core area has seen a significant decline in new patent filings, dropping by 30.8% in 2025 and 77.8% so far in 2026. While Machine Learning & AI and Web & Cloud Service Protocols showed rapid growth in 2024 and 2025 (100.0% and 300.0% YoY respectively), the sharp decline across nearly all categories so far in 2026, including a 100.0% drop in Network Management & Monitoring, suggests a dramatic shift or slowdown in its patenting activity.

Patent Trend by Technology Area

Yearly patent publications since 2023

Product themes

Product-level themes inferred from filings since 2023, with category chips showing where each theme appears. Select a theme to filter the patents below.

47 US filings (since 2023) · 12 categories · 10 themes

Network Operations, Reliability & Traffic

Techniques and systems designed to monitor network health, diagnose issues, optimize traffic flow, and ensure continuous operation and reduced downtime in complex network environments, including cloud and storage area networks.

Routing, Switching & QoSNetwork Management & Monitoring
Who else files here? →
23since 2023
-91.7%YoY
Network Traffic Management & QoS

Techniques and systems for optimizing network traffic flow, distributing loads across multiple paths or resources, and ensuring quality of service based on various criteria like application type, latency, or resource availability. This includes dynamic path selection, congestion control, and resource allocation.

Routing, Switching & QoS
Who else files here? →
21since 2023
-77.8%YoY
Hardware Network Acceleration

Utilizing specialized hardware components such as network processing units (NPUs) or dedicated acceleration circuits to offload and speed up network packet processing, traffic generation, or time-sensitive network operations.

Network Management & Monitoring
Who else files here? →
13since 2023
+25.0%YoY
Distributed Cloud/Edge Processingfiltered

Techniques for distributing computational tasks, data storage, and service logic across cloud data centers, edge devices, and user equipment to improve performance, resilience, or resource utilization. This includes architectures for split rendering, decentralized ledgers, and microservices.

Web & Cloud Service Protocols
Who else files here? →
7since 2023
+100.0%YoY
Automated Network Provisioning

Systems and methods for automatically deploying, configuring, and updating network devices and services, including software updates, client onboarding, and topology management across various network types.

Network Management & Monitoring
Who else files here? →
5since 2023
new
Physical Layer & Interface Optimization

Enhancements to the physical and data link layers of network communication, focusing on hardware components, signal integrity, power efficiency, and efficient data transfer mechanisms for specific interfaces and buses.

Routing, Switching & QoS
Who else files here? →
3since 2023
new
Network Data Management

Systems and methods for collecting, processing, and ensuring the quality and consistency of data used for network monitoring, asset management, and operational decision-making, including conflict detection and reliability scoring.

Network Management & Monitoring
Who else files here? →
3since 2023
0.0%YoY
3D Die Stacking & Vertical Interconnects

Techniques for stacking multiple semiconductor dies or active layers vertically to achieve higher density and shorter interconnections, often utilizing through-silicon vias (TSVs) or other vertical conductive paths like through-hole electrodes.

Semiconductor Packaging & Encapsulation
Who else files here? →
1since 2023
new
Advanced Package Interconnects

Novel electrical connection methods within or between package components, including through-glass vias (TGVs), backside contacts, and optimized redistribution layers for improved signal and power integrity.

Semiconductor Packaging & Encapsulation
Who else files here? →
1since 2023
new
Cable Termination & Management

Systems and methods for preparing, organizing, and electrically connecting individual conductors or cable bundles to connector terminals, ensuring reliable contact and strain relief.

Electrical Connectors
Who else files here? →
1since 2023
n/a

Patents

Showing 11-15 of 15

Distributed Cloud/Edge Processing
Page 2 of 2
US 11328222 B1GRANTED
G06F15/18

Network switch with integrated gradient aggregation for distributed machine learning

Filed:2019-05-10Pub:2022-05-10
Applicant:Innovium, Inc.

Distributed machine learning systems and other distributed computing systems are improved by embedding compute logic at the network switch level to perform collective actions, such as reduction operations, on gradients or other data processed by the nodes of the system. The switch is configured to recognize data units that carry data associated with a collective action that needs to be performed by the distributed system, referred to herein as “compute data,” and process that data using a compute subsystem within the switch. The compute subsystem includes a compute engine that is configured to perform various operations on the compute data, such as “reduction” operations, and forward the results back to the compute nodes. The reduction operations may include, for instance, summation, averaging, bitwise operations, and so forth. In this manner, the network switch may take over some or all of the processing of the distributed system during the collective phase.

US 11099902 B1GRANTED
G06F9/50

Parallelized ingress compute architecture for network switches in distributed artificial intelligence and other applications

Filed:2019-05-10Pub:2021-08-24
Applicant:Innovium, Inc.

Distributed machine learning systems and other distributed computing systems are improved by embedding compute logic at the network switch level to perform collective actions, such as reduction operations, on gradients or other data processed by the nodes of the system. The switch is configured to recognize data units that carry data associated with a collective action that needs to be performed by the distributed system, referred to herein as “compute data,” and process that data using a compute subsystem within the switch. The compute subsystem includes a compute engine that is configured to perform various operations on the compute data, such as “reduction” operations, and forward the results back to the compute nodes. The reduction operations may include, for instance, summation, averaging, bitwise operations, and so forth. In this manner, the network switch may take over some or all of the processing of the distributed system during the collective phase.

US 11057318 B1GRANTED
H04L12/935

Distributed artificial intelligence extension modules for network switches

Filed:2019-08-27Pub:2021-07-06
Applicant:Innovium, Inc.

Distributed machine learning systems and other distributed computing systems are improved by compute logic embedded in extension modules coupled directly to network switches. The compute logic performs collective actions, such as reduction operations, on gradients or other compute data processed by the nodes of the system. The reduction operations may include, for instance, summation, averaging, bitwise operations, and so forth. In this manner, the extension modules may take over some or all of the processing of the distributed system during the collective phase. An inline version of the module sits between a switch and the network. Data units carrying compute data are intercepted and processed using the compute logic, while other data units pass through the module transparently to or from the switch. Multiple modules may be connected to the switch, each coupled to a different group of nodes, and sharing intermediate results. A sidecar version is also described.

US 10931602 B1GRANTED
H04L12/931

Egress-based compute architecture for network switches in distributed artificial intelligence and other applications

Filed:2019-05-10Pub:2021-02-23
Applicant:Innovium, Inc.

Distributed machine learning systems and other distributed computing systems are improved by embedding compute logic at the network switch level to perform collective actions, such as reduction operations, on gradients or other data processed by the nodes of the system. The switch is configured to recognize data units that carry data associated with a collective action that needs to be performed by the distributed system, referred to herein as “compute data,” and process that data using a compute subsystem within the switch. The compute subsystem includes a compute engine that is configured to perform various operations on the compute data, such as “reduction” operations, and forward the results back to the compute nodes. The reduction operations may include, for instance, summation, averaging, bitwise operations, and so forth. In this manner, the network switch may take over some or all of the processing of the distributed system during the collective phase.

US 10931588 B1GRANTED
H04L12/24

Network switch with integrated compute subsystem for distributed artificial intelligence and other applications

Filed:2019-05-10Pub:2021-02-23
Applicant:Innovium, Inc.

Distributed machine learning systems and other distributed computing systems are improved by embedding compute logic at the network switch level to perform collective actions, such as reduction operations, on gradients or other data processed by the nodes of the system. The switch is configured to recognize data units that carry data associated with a collective action that needs to be performed by the distributed system, referred to herein as “compute data,” and process that data using a compute subsystem within the switch. The compute subsystem includes a compute engine that is configured to perform various operations on the compute data, such as “reduction” operations, and forward the results back to the compute nodes. The reduction operations may include, for instance, summation, averaging, bitwise operations, and so forth. In this manner, the network switch may take over some or all of the processing of the distributed system during the collective phase.

12