DCT

3:25-cv-03738

Netflix Inc v. Broadcom Inc

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 3:25-cv-03738, N.D. Cal., 05/22/2025
  • Venue Allegations: Venue is alleged to be proper as both Defendants maintain principal places of business within the Northern District of California and have committed acts of patent infringement in the district.
  • Core Dispute: Plaintiff alleges that Defendants’ cloud virtualization and networking products infringe two patents related to virtual machine management and subnet provisioning, and that Defendants’ Ethernet switching hardware and software infringe three patents related to high-precision network time synchronization.
  • Technical Context: The technologies at issue relate to foundational elements of modern cloud computing infrastructure and high-performance data center hardware, addressing challenges in service availability, network configuration, and time-sensitive data processing.
  • Key Procedural History: The complaint follows Broadcom's 2023 acquisition of VMware, with Plaintiff alleging that Broadcom is now liable for infringement by legacy VMware products. Plaintiff alleges it provided Defendants with notice of infringement for the asserted patents via letters dated December 2024 and April 2025. The complaint also cites prior litigation where juries found VMware liable for willful infringement, alleging a "culture of copying" that it claims Broadcom continues.

Case Timeline

Date Event
2003-03-17 ’102 Patent Priority Date
2005-04-27 ’912 Patent Priority Date
2005-12-09 ’931 and ’751 Patents Priority Date
2007-12-25 ’102 Patent Issue Date
2008-11-04 ’931 Patent Issue Date
2010-01-19 ’912 Patent Issue Date
2010-02-02 ’751 Patent Issue Date
2014-08-29 ’472 Patent Priority Date
2019-06-25 ’472 Patent Issue Date
2023-11-22 Broadcom completes acquisition of VMware
2024-12-23 Netflix sends first notice letter to Defendants
2025-04-15 Netflix sends second notice letter to Defendants
2025-05-22 Complaint Filing Date

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 10,331,472, "Virtual Machine Service Availability," issued June 25, 2019

The Invention Explained

  • Problem Addressed: The patent describes the challenge of ensuring high availability for services in virtualized environments. Conventional methods using redundant standby services were noted to be costly and could leave a critical service vulnerable if a primary server failed, creating a "risky period of lower service availability" until the server was restored (’472 Patent, col. 1:44-61).
  • The Patented Solution: The invention proposes a dynamic method for managing service availability. A controller monitors services running on virtual machines. If a high-priority service's availability drops, the system identifies a lower-priority service that has excess availability, deactivates it to free up server resources, and then activates a new instance of the high-priority service on the freed server. This approach claims to improve reliability and resource utilization without requiring permanently dedicated hardware for every critical service (’472 Patent, col. 1:62-2:8).
  • Technical Importance: This method of dynamic resource reallocation sought to improve the cost-efficiency and resilience of virtualized network functions, a key objective in the growth of cloud computing infrastructure (Compl. ¶25).

Key Claims at a Glance

  • The complaint asserts claims 6-10, with independent claim 6 being central (Compl. ¶100).
  • Independent Claim 6 recites a method with the essential elements:
    • Monitoring a first availability of a first service, which has a first availability requirement and tolerance.
    • Detecting a reduction in the first service's availability.
    • Creating capacity by deactivating a second service on an active virtual machine, where the second service has availability exceeding its tolerance and a lower availability requirement than the first service.
    • Activating a second active virtual machine on the server to execute the first service.

U.S. Patent No. 7,313,102, "System and Method for Subnet Configuration and Selection," issued December 25, 2007

The Invention Explained

  • Problem Addressed: The patent identifies inefficiencies in prior art network subnet management. These tools were described as managing only IP address space and lacking the ability to handle related characteristics like performance and security, requiring manual coordination between network managers and consumers, which could lead to delays and errors (’102 Patent, col. 1:29-48).
  • The Patented Solution: The patent discloses a method for simplifying subnet provisioning. The solution involves first grouping subnets based on "logical properties" (e.g., security, performance) and assigning authorized groups to specific network consumers. It then provides a graphical user interface that guides the consumer through a "constrained selection" process, presenting only valid and accessible options for parameters like address type, gateway, and subnet mask. This is intended to streamline configuration and prevent errors (’102 Patent, Abstract; col. 1:55-2:13).
  • Technical Importance: This invention aimed to automate and simplify a critical network administration task, improving the accuracy, security, and efficiency of resource allocation in complex enterprise networks (Compl. ¶38).

Key Claims at a Glance

  • The complaint asserts claims 1-11, with independent claim 1 being central (Compl. ¶151).
  • Independent Claim 1 recites a method with the essential elements:
    • Grouping subnets into subnet groups based on logical properties.
    • Assigning accessible subnet groups to each network consumer.
    • Providing for constrained selection of a subnet via a graphical user interface with selectable fields.
    • The constrained selection includes steps for selecting (i) a public or private address space, (ii) an applicable gateway device, (iii) a subnet group, and (iv) a subnet mask representing the size.

U.S. Patent No. 7,649,912, "Time Synchronization, Deterministic Data Delivery and Redundancy for Cascaded Nodes on Full Duplex Ethernet Networks," issued January 19, 2010

  • Technology Synopsis: The patent addresses challenges in achieving high-precision time synchronization in daisy-chained Ethernet networks, which are common in industrial automation. The invention discloses a method for intermediate nodes (e.g., network switches) to correct for internal processing delays by adjusting timestamps in IEEE 1588 synchronization messages "on the fly," thereby improving synchronization accuracy across the network (Compl. ¶¶45-48).
  • Asserted Claims: Claims 1-3 and 5-12 (Compl. ¶193).
  • Accused Features: Broadcom's ethernet switching products (e.g., StrataDNX devices), firmware, and software (e.g., BroadPTP 1588 Software Suite) that implement the IEEE 1588 Precision Time Protocol (PTP) (Compl. ¶193).

U.S. Patent No. 7,447,931, "Step time change compensation in an industrial automation network," issued November 4, 2008

  • Technology Synopsis: The patent targets the "step change problem," where an abrupt change in a master clock's time can cause slave clocks to become desynchronized, potentially disrupting time-sensitive industrial processes. The disclosed method enables a node to detect such a step change by receiving and comparing successive timestamps and their associated offsets from a source, and then selectively updating its local timestamp to compensate for the change (Compl. ¶¶64-65, 72).
  • Asserted Claims: Claims 27-32 (Compl. ¶241).
  • Accused Features: Broadcom's ethernet switching products, firmware, and software that implement IEEE 1588 PTP, which allegedly perform the claimed step change compensation (Compl. ¶241).

U.S. Patent No. 7,656,751, "Step time change compensation in an industrial automation network," issued February 2, 2010

  • Technology Synopsis: Sharing a common specification with the ’931 Patent, this patent claims a system for addressing the same "step change problem." It describes a system with a "timestamp component" for capturing timestamps and offsets and a "time synch component" that identifies step changes from a master clock and synchronizes the local clock time accordingly (Compl. ¶¶78-79, 86).
  • Asserted Claims: Claims 1-14 (Compl. ¶306).
  • Accused Features: Broadcom's ethernet switching products, firmware, and software that implement IEEE 1588 PTP, which allegedly contain the claimed system for step change compensation (Compl. ¶306).

III. The Accused Instrumentality

Product Identification

The complaint names two categories of accused products:

  1. Broadcom Load Balancing Accused Products & Broadcom Subnet Provisioning Accused Products: These are virtualization and cloud networking platforms acquired from VMware, including VMware Cloud Foundation, VMware Cloud on AWS, and products incorporating VMware NSX/NSX-T Data Center and the VMware Avi Load Balancer (Compl. ¶¶100, 151).
  2. Broadcom Switching Accused Products: These are Broadcom's hardware and software products for ethernet switching, including StrataDNX switch/router devices, BroadPTP 1588 Software Suite, and BroadSync firmware (Compl. ¶¶193, 241, 306).

Functionality and Market Context

  • The VMware/NSX-based products provide the core infrastructure for modern private and hybrid cloud environments, enabling software-defined networking, security, and load balancing. The complaint alleges that the VMware Avi Load Balancer uses "Service Engines," described as data plane virtual machines, to host virtual services and automatically scales these services based on performance metrics (Compl. ¶¶104, 107). It further alleges that the NSX platform provides a "Virtual Private Cloud (VPC)" abstraction layer that allows users to provision and configure network subnets through a graphical interface (Compl. ¶154).
  • The Broadcom Switching products are high-performance hardware components and related software used in data centers, automotive, and industrial networks. The complaint alleges these products implement the IEEE 1588 Precision Time Protocol (PTP) to enable highly accurate time synchronization between networked devices, a critical feature for applications requiring deterministic data delivery (Compl. ¶¶193, 196, 244). The complaint provides a visual from Broadcom's product briefs highlighting hardware support for "IEEE 1588v2 and SyncE implementations with nanosecond-scale time stamping" (Compl. ¶193, Fig. 30).

IV. Analysis of Infringement Allegations

10,331,472 Infringement Allegations

Claim Element (from Independent Claim 6) Alleged Infringing Functionality Complaint Citation Patent Citation
A method comprising: monitoring a first availability of a first service, the first service having a first availability requirement and a first availability tolerance; The accused products monitor "virtual services" and display a numeric, color-coded "health" status. They also use settings for "minimum and maximum scale-out per virtual service" which allegedly govern availability requirements and tolerances. ¶102, ¶103 col. 9:18-21
detecting a reduction in the first availability of the first service; The products use metric-based thresholds, such as detecting when a Service Engine's (SE) CPU usage "exceeds an 80% average," to trigger a scale-out or migration operation, which constitutes detecting reduced availability. This is depicted in a flowchart from the product's documentation (Compl. ¶109, Fig. 7). ¶105, ¶108 col. 9:22-23
creating capacity for the first service by deactivating a second service on a first active virtual machine on a server, the second service having a second availability exceeding a second availability tolerance and having a second availability requirement lower than the first availability requirement; In a "migrate" operation, a new SE is added to a virtual service, made primary, and after a 30-second transition, the "old SE will terminate the remaining connections and be removed from the virtual service configuration." This removal of the old SE is alleged to be the "deactivating" step that creates capacity. ¶109, ¶110 col. 9:24-31
and activating a second active virtual machine executing the first service on the server. In a "scale out" operation, the system adds "one or more new SEs to the virtual service." An SE is identified as a virtual machine, and adding a new one is alleged to be activating a second active virtual machine to execute the service. ¶112, ¶113 col. 9:32-34

Identified Points of Contention (’472 Patent):

  • Scope Questions: A primary dispute may arise over whether the accused "migrate" function meets the "deactivating a second service" limitation. Defendants may argue that migrating a single service by adding a new instance and removing the old is a handoff process, not the deactivation of a distinct, lower-priority service to "create capacity" as described in the patent's problem statement.
  • Technical Questions: The complaint alleges CPU usage as the trigger for detecting reduced availability. A key question will be whether Plaintiff can provide evidence that the accused products also evaluate the specific availability "requirements" and "tolerances" of different services as required by the claim's "deactivating" step, a point on which the complaint provides limited detail. The complaint's visual evidence, an annotated diagram of a "Shared Responsibility Model," is used to allege VMware's control over the "NSX Lifecycle" (Compl. ¶118, Fig. 11).

7,313,102 Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
A method for provisioning subnets, the method comprising: grouping the subnets into subnet groups based on logical properties of the subnets; The accused NSX products provide for selecting from two groups of subnets, "tier-0" and "tier-1," which are alleged to be groups based on the logical property of routing architecture (e.g., external vs. internal connectivity). Screenshots from the NSX Administrator Guide show these distinct source types (Compl. ¶154, Fig. 17 & 18). ¶153, ¶154 col. 9:15-17
assigning to each network consumer those subnet groups that are accessible to that network consumer; The NSX platform allows administrators to assign users to a Virtual Private Cloud (VPC), and those users can then add subnets within their assigned VPC. This is alleged to constitute assigning accessible subnet groups. The complaint includes a screenshot from product documentation stating users can add subnets "inside the NSX VPC that is assigned to them" (Compl. ¶157, Fig. 19). ¶156, ¶157 col. 9:18-20
and providing for constrained selection of a particular subnet by a network consumer accomplished by way of a graphical user interface with selectable fields, wherein the constrained selection includes (i) selecting a public or private type address space... and (iv) selecting a subnet mask... The accused products provide a GUI for adding a subnet where a user specifies properties. This allegedly includes (i) selecting an "Access Mode" of "Private, Public, Isolated," as shown in a product guide screenshot (Compl. ¶160, Fig. 20), (ii) specifying a "Gateway IP" (Compl. ¶161, Fig. 21), and (iv) selecting a "Size" for the subnet, which represents the subnet mask (Compl. ¶160, Fig. 20). ¶158, ¶160 col. 9:21-29

Identified Points of Contention (’102 Patent):

  • Scope Questions: A likely point of contention is whether the architectural "tier-0" and "tier-1" routing constructs in NSX constitute "subnet groups based on logical properties" as contemplated by the patent. Defendants may argue the patent envisions more flexible, user-defined logical groupings (e.g., by department or security level) rather than fixed routing tiers inherent to the product's design.
  • Technical Questions: What evidence does the complaint provide that a user performs the explicit step of "selecting a subnet group from those subnet groups that are accessible," as required by element (iii) of the "constrained selection" limitation? The complaint's theory appears to be that adding a subnet within an assigned VPC implicitly makes this selection, which may not map directly to the claim's language.

V. Key Claim Terms for Construction

’472 Patent

  • The Term: "deactivating a second service"
  • Context and Importance: This term is critical because the Plaintiff's infringement theory for the "migrate" functionality hinges on construing the removal of an old virtual machine instance (after a new one is activated) as "deactivating a second service." A narrow construction requiring the shutdown of a distinct, lower-priority service could undermine this theory.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The patent states that "an instance of a lower priority service is deactivated to provide an available server" (’472 Patent, col. 9:4-6). Plaintiff may argue "an instance of a... service" is itself a "service" that can be deactivated, allowing the term to cover the removal of the old SE instance.
    • Evidence for a Narrower Interpretation: The patent's background distinguishes between services with different functional roles and priority levels, such as a "business and database layer" versus a "management layer" (’472 Patent, col. 1:39-44). This context suggests "a second service" may refer to a functionally separate service, not merely another instance of the first service.

’102 Patent

  • The Term: "subnet groups based on logical properties"
  • Context and Importance: The infringement case depends on whether NSX's "tier-0" and "tier-1" routing structures qualify as the claimed "subnet groups." Practitioners may focus on this term because if these architectural tiers are not "groups" based on "logical properties," the entire infringement allegation for claim 1 may fail.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification explicitly lists "route information" as an example of a "logical propert[y]" that can be used for grouping (’102 Patent, col. 3:41-42). Plaintiff will likely argue that tier-0 (external connectivity) and tier-1 (internal connectivity) are defined by their route information.
    • Evidence for a Narrower Interpretation: The patent also provides examples like "security characteristic" and "performance characteristic" and discusses assigning groups based on "workgroup types" like "human resources, finance, administration, and engineering" (’102 Patent, col. 3:35-42, col. 3:48-56). This may suggest the invention was directed at more abstract, policy-based groupings rather than fundamental routing architecture.

VI. Other Allegations

Indirect Infringement

For all asserted patents, the complaint alleges both induced and contributory infringement. The inducement allegations are based on Defendants allegedly providing instructional materials, user guides, marketing, technical support, and interactive "Hands-on Labs" that instruct and encourage customers to use the accused products in an infringing manner (Compl. ¶¶126-134, 172-179, 226-228, 291-293, 337-339). Contributory infringement is alleged on the basis that the accused products contain components specially made to practice the claimed inventions and are not staple articles of commerce suitable for substantial non-infringing use when configured to perform the accused functions (Compl. ¶¶137-140, 182-185, 230-233, 295-298, 341-344).

Willful Infringement

Willfulness is alleged for all five patents. The claims are primarily based on pre-suit knowledge from notice letters sent in December 2024 and April 2025 (Compl. ¶¶143, 188, 236, 301, 347). The complaint further supports its willfulness claims by alleging that VMware has a "demonstrated culture of knowingly using patented technology" and a "culture of copying," citing adverse jury verdicts for willful infringement in prior, unrelated litigation (Compl. ¶¶146, 148).

VII. Analyst’s Conclusion: Key Questions for the Case

  • A core issue will be one of functional equivalence: does the accused "migration" process in the VMware Avi Load Balancer, which involves a handoff between two instances of the same service, perform the same function in substantially the same way as the ’472 Patent's claimed step of "deactivating a second service" to "creat[e] capacity"? The answer may depend on whether the claim requires shutting down a functionally distinct, lower-priority service.
  • A central dispute will be one of technical categorization: are the accused NSX platform's fundamental "tier-0" and "tier-1" routing layers, which are based on network architecture, the same as the ’102 Patent's "subnet groups based on logical properties," which the specification suggests can be based on abstract business policies like department or security level?
  • A key question for damages will be one of corporate conduct: will Plaintiff's allegations of a "culture of copying," supported by prior willful infringement verdicts against VMware, be deemed relevant and persuasive in establishing willful infringement by both VMware and its successor, Broadcom, in this case?