DCT

5:26-cv-02379

Valtrus Innovations Ltd v. Google LLC

Log In
Key Events
Amended Complaint
complaint Intelligence

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 5:26-cv-02379, N.D. Cal., 05/21/2026
  • Venue Allegations: Venue is alleged to be proper in the Northern District of California because Google maintains its corporate headquarters and multiple other regular and established places of business in the district, and has committed acts of infringement there.
  • Core Dispute: Plaintiff alleges that Defendant's Google Search, Google Cloud, and related services infringe three patents related to methods for merging search results and for hardware-based processor utilization metering.
  • Technical Context: The technologies at issue concern foundational aspects of modern large-scale computing: efficiently combining results from disparate data sources in web search, and accurately metering CPU usage in partitioned, multi-tenant server environments.
  • Key Procedural History: The complaint references a prior lawsuit filed by Plaintiff against Google on January 10, 2022 (referred to as Valtrus I), asserting infringement of the same patents. For U.S. Patent No. 6,816,809, the complaint notes that the Patent Trial and Appeal Board denied institution of an inter partes review (IPR) petition filed by Google, and that in a subsequent ex parte reexamination (EPR), certain claims were canceled while others, including asserted claims 2-5, 11, and 13-17, were confirmed as valid. Plaintiff also notes that prosecution history for the '704 patent involved overcoming rejections by adding specific claim limitations.

Case Timeline

Date Event
1999-10-15 '604 Patent Priority Date
2001-08-27 '704 Patent Priority Date
2002-07-23 '809 Patent Priority Date
2004-04-27 '704 Patent Issued
2004-11-09 '809 Patent Issued
2007-05-16 Google launches "Universal Search" feature (approximate date based on cited blog post)
2008-03-18 '604 Patent Issued
2012-06-26 Alleged knowledge date for '604 Patent (via citing patent)
2013-01-01 Alleged knowledge date for '704 Patent (via citing patent)
2021-04-14 Alleged knowledge date for '809 Patent (via notice letter)
2022-01-10 Prior lawsuit (Valtrus I) filed against Google
2022-07-05 Google files IPR petition on '809 Patent
2023-01-03 PTAB denies institution of IPR on '809 Patent
2024-02-21 Google files for EPR on '809 Patent
2026-05-21 Complaint Filing Date

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

U.S. Patent No. 6,728,704 - "Method and apparatus for merging result lists from multiple search engines"

  • Patent Identification: U.S. Patent No. 6,728,704, "Method and apparatus for merging result lists from multiple search engines," issued April 27, 2004.

The Invention Explained

  • Problem Addressed: The patent's background describes the problem of merging multiple search result lists into a single, ranked list Compl. ¶17 Prior methods were computationally expensive because they required examining and ranking every single entry from every list, which could "nullify any advantage gained by operating multiple search engines at the same time" for large data sets Compl. ¶17 '704 Patent, col. 2:48-56
  • The Patented Solution: The invention claims to reduce this computational overhead by proposing a multi-step method. Instead of processing every result, the method first selects a smaller "subset" of entries from each result list Compl. ¶21 '704 Patent, col. 5:44-48 It then assigns a "representative value" to each list based on the scores of the entries within its subset Compl. ¶21 '704 Patent, col. 8:10-13 Finally, it produces a merged list by ranking the result lists themselves based on their representative values, which can vary dynamically during the merging process Compl. ¶¶22-24 '704 Patent, abstract
  • Technical Importance: This approach enabled search engines to more efficiently aggregate and rank information from numerous distinct sources, improving response time without sacrificing the quality of the merged results Compl. ¶19

Key Claims at a Glance

  • The complaint asserts independent claim 1 and system claim 12, along with dependent claims 2-6 and 13-17 Compl. ¶34 Compl. ¶¶62-77 Compl. ¶78
  • Independent Claim 1 requires the steps of:
    • transmitting a query to a set of search engines;
    • receiving a result list from each search engine;
    • selecting a subset of entries from each result list;
    • assigning a scoring value to each entry in the subset;
    • assigning to each subset a representative value based on those scoring values;
    • producing a merged list of entries based on the representative values;
    • wherein the representative value varies in a predetermined manner.
  • The complaint reserves the right to assert additional claims Compl. ¶34

U.S. Patent No. 6,816,809 - "Hardware based utilization metering"

  • Patent Identification: U.S. Patent No. 6,816,809, "Hardware based utilization metering," issued November 9, 2004.

The Invention Explained

  • Problem Addressed: The patent addresses the challenges of measuring CPU utilization in computer systems where hardware can be partitioned to run multiple, isolated operating systems Compl. ¶132 Software-based metering in such environments is problematic because it requires communication between a metering application and each isolated OS, which is not a default capability and can be disrupted by network or OS failures (Compl. ¶132; Compl. ¶133, Compl. ¶¶col. 2:50-66).
  • The Patented Solution: The patent proposes a hardware-based solution that operates independently of the operating systems. The core components are a hardware "idle indicator" that detects when a processor is busy, a hardware "counter" that measures the amount of time the processor is busy based on the indicator and a system clock, and a "data usage provider" that can store and report the counter's value, even maintaining it in non-volatile memory Compl. ¶134 '809 Patent, abstract '809 Patent, col. 4:1-23
  • Technical Importance: This method provides a reliable and accurate way to meter CPU usage, which is a critical function for billing and resource management in pay-per-use, multi-tenant computing environments like cloud platforms and server farms Compl. ¶131 '809 Patent, col. 2:44-48

Key Claims at a Glance

  • The complaint asserts independent method claim 13 and apparatus claim 18, along with dependent claims 2-5, 11, and 14-20 Compl. ¶140 Compl. ¶147 Compl. ¶162 Compl. ¶165 Compl. ¶168 Compl. ¶172 Compl. ¶176 Compl. ¶200 Compl. ¶203 Compl. ¶206 Compl. ¶209
  • Independent Claim 13 (a method) requires the steps of:
    • determining when any of a plurality of processors is busy;
    • providing a busy indication to a counter;
    • receiving at the counter a measure of computer system time;
    • incrementing a counter value based on the busy indication and the system time;
    • maintaining the counter value.
  • Independent Claim 18 (an apparatus) recites means for performing analogous functions for each of a plurality of CPUs arranged in cells.

U.S. Patent No. 7,346,604 - "Method for ranking hypertext search results by analysis of hyperlinks from expert documents and keyword scope"

  • Patent Identification: U.S. Patent No. 7,346,604, "Method for ranking hypertext search results by analysis of hyperlinks from expert documents and keyword scope," issued March 18, 2008.

Technology Synopsis

  • The patent addresses the difficulty of ranking broad search queries and filtering out "spam" pages created to mislead search engines Compl. ¶253 The patented "Hilltop" method provides a two-phase solution: first, it identifies a set of "expert documents" from the web in a pre-processing step, independent of any specific query Compl. ¶257 Compl. ¶276 Second, upon receiving a user query, it ranks those expert documents and then ranks the "target documents" to which the experts link, using this analysis to generate a final, more authoritative list of results Compl. ¶260 '604 Patent, abstract

Asserted Claims

  • Independent claim 1 is asserted, along with dependent claims 2-20 Compl. ¶268 Compl. ¶¶295-369
  • The complaint alleges that Google Search, Google News, and related internal systems infringe by practicing the "Hilltop" algorithm to identify expert pages and rank results based on links from those experts Compl. ¶250 Compl. ¶276 Compl. ¶290

III. The Accused Instrumentality

  • Product Identification: The accused instrumentalities are Google Search and Google Cloud products Compl. ¶34 Compl. ¶140
  • Functionality and Market Context:
    • Google Search: The complaint alleges Google Search is the dominant U.S. search engine Compl. ¶6 The accused functionality involves its "Universal Search" feature, which integrates results from various specialized internal data sources-such as Web, Maps, News, Videos, and Images-into a single, merged search results page (SERP) Compl. ¶32 Compl. ¶36 The complaint specifically names internal Google systems like "Superroot," "Tangram," and "Universal Packer" as being involved in this process of combining results from multiple corpora Compl. ¶36 The complaint includes a screenshot of a search for "in-n-out burger" to illustrate how results from different sources (web links, a map, and news stories) are combined into one list Compl. ¶36
    • Google Cloud: The complaint accuses Google Cloud products, which provide virtual machines (VMs) for customers Compl. ¶141 Specifically, it identifies C3D, N2D, T2D, and C2D VMs that run on AMD EPYC processors Compl. ¶141 The complaint alleges these products and internal Google systems like "Borg" use hardware counters for CPU utilization to enable features like predictive autoscaling and to manage power consumption Compl. ¶¶143-146 The complaint provides a marketing chart showing AMD EPYC powered machine families on Google Cloud Compl. ¶141

IV. Analysis of Infringement Allegations

'704 Patent Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
transmitting a query to a set of search engines A user's query is transmitted to multiple Google search services, such as Web, Maps, and News. ¶42 col. 8:2-3
receiving in response to said query a result list from each search engine of said set of search engines Google Search receives result lists from each of its distinct search services in response to the query. ¶45 col. 8:4-6
selecting a subset of entries from each result list to form a set of selected entries Google Search returns millions of results but selects only a subset of those entries for display to the user. ¶48 col. 8:7-8
assigning to each selected entry of said set of selected entries a scoring value according to a scoring function Entries within each result list are sorted by relevance (e.g., by distance for Maps results), which is alleged to be based on an assigned scoring value. ¶51 col. 8:9-11
assigning to each subset a representative value according to the scoring values assigned to its entries Google Search allegedly orders the groups of results (e.g., Maps vs. News) based on a representative value assigned to each group. ¶54 col. 8:12-13
producing a merged list of entries in a predetermined manner based on the representative value assigned to each result list The final SERP is a merged list where the order of result groups (e.g., Maps appearing before News) is based on their assigned representative value. ¶58 col. 8:14-16
wherein the representative value varies in accordance with predetermined manner The ranking of result groups changes depending on the query (e.g., Maps results are higher for location-based queries), suggesting the representative value varies. ¶60 col. 8:17-18

'809 Patent Infringement Allegations

Claim Element (from Independent Claim 13) Alleged Infringing Functionality Complaint Citation Patent Citation
determining when any of the plurality of processors is busy Processors used in Google Cloud have performance counters (e.g., MPERF, APERF) that are used to determine when a processor core is in a C0 (busy) state. ¶182 col. 8:16-17
providing a busy indication to a counter associated with a busy processor The processor core provides a busy indication to a counter while the core is in the C0 state, which controls when the counter increments. ¶185 col. 8:18-20
receiving at the counter a measure of computer system time The APERF counter in the accused processors receives a core clock as a pulse train, which is a measure of system time. ¶188 col. 8:21-22
incrementing a counter value in the counter based on the provided busy indication and an amount of computer system time that the processor is determined to be busy The MPERF and APERF counters are incremented by hardware based on the busy indication (C0 state) and clock cycles. ¶192 col. 8:23-27
maintaining the counter value The counters in the accused processors are read/write registers, meaning their values are preserved for access by software. ¶195 col. 8:28-29

Identified Points of Contention

  • '704 Patent: The dispute may center on definitional scope. A primary question is whether Google's various internal data services (e.g., Maps, News, Images) constitute distinct "search engines" as contemplated by the patent, or if they are simply different data verticals within a single, monolithic search architecture. Another point of contention may be whether Google's system of ranking blocks of results on a SERP is technically equivalent to the patent's claimed process of assigning a "representative value" to a "subset" of results and merging individual entries based on that value.
  • '809 Patent: The core of the dispute may be one of technical implementation and claim scope. A key question is whether the use of standard, multi-purpose hardware performance counters (such as MPERF and APERF in AMD EPYC processors) for functions like autoscaling satisfies the specific "hardware based utilization metering" limitations. The defense might argue that the patent claims a dedicated, novel circuit arrangement, not the use of pre-existing, general-purpose counters. The complaint's reliance on redacted information from a prior case suggests that the public-facing evidence of how these counters operate may be a central battleground.

V. Key Claim Terms for Construction

For the '704 Patent

  • The Term: "search engine"
  • Context and Importance: The infringement theory requires merging results from multiple "search engines" Compl. ¶35 The viability of the plaintiff's case depends on whether Google's internal, specialized data sources (Maps, News, etc.) are construed as separate "search engines". Practitioners may focus on this term because if Google Search is deemed a single, integrated engine querying internal databases, the "multiple search engines" limitation may not be met.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification discusses "federated, peer-to-peer, and meta-search engines" '704 Patent, col. 1:53-62 and describes search engines as computer programs that can be resident on different computers in a network (e.g., computer 10 vs. 30) '704 Patent, col. 3:24-44 This may support an interpretation where functionally distinct, networked software components qualify as separate "search engines", even if owned by the same entity.
    • Evidence for a Narrower Interpretation: The background discusses farming out a search "to other public web search engines" '704 Patent, col. 2:32-33 This language could support an argument that a "search engine" refers to a separate, independent, and likely third-party entity, not an internal component of a single company's infrastructure.

For the '809 Patent

  • The Term: "idle indicator"
  • Context and Importance: The claims require a specific hardware-based "idle indicator" coupled to the processor '809 Patent, claim 1 The infringement case hinges on whether a standard processor's state-reporting mechanism, such as the C0 "busy" state that performance counters read, qualifies as this claimed "idle indicator". Practitioners may focus on this term because the patent's novelty was predicated on its hardware-based approach to distinguish it from prior art software solutions Compl. ¶138
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The claim language is functional, reciting "an idle indicator coupled to a processor, wherein the idle indicator receives an indication when the processor is in a first state" '809 Patent, claim 1 This could support a broad interpretation covering any hardware mechanism that performs this function, including standard processor state flags.
    • Evidence for a Narrower Interpretation: The specification describes specific embodiments, including a "hardware modification to the computer system" and a device "coupled to the pin to read the halt (idle) indication" '809 Patent, col. 4:36-51 This language, along with prosecution history arguments distinguishing the invention from software, may support a narrower construction requiring a specific, added hardware component rather than just the inherent state-reporting functionality of a standard CPU.

VI. Other Allegations

  • Indirect Infringement: The complaint alleges that Google induces infringement of all asserted patents. For the '704 and '604 patents, this is based on offering Google Search to the public with the intent to encourage infringing use Compl. ¶124 Compl. ¶372 For the '809 patent, this is based on offering Google Cloud products that rely on processors with the claimed metering hardware, with the intent to encourage infringing use Compl. ¶242
  • Willful Infringement: Willfulness is alleged for all three patents. The allegations are based on both pre-suit and post-suit knowledge. The complaint alleges pre-suit knowledge stemming from: (1) Google's own patents citing the patents-in-suit as references; (2) Google's employment of certain co-inventors; (3) a notice letter sent for the '809 patent; and (4) a prior lawsuit filed on January 10, 2022, concerning the same patents Compl. ¶¶122-123 Compl. ¶¶240-241 Compl. ¶¶370-371

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

This case presents several fundamental questions for the court, blending issues of claim construction with the technical realities of large-scale cloud and search infrastructure.

  • A core issue will be one of definitional scope: Can the term "search engine", as used in the '704 patent, be construed to cover the various internal data silos (Maps, News, Video) of a single, integrated search provider like Google? The outcome of this question may determine whether the foundational "multiple search engines" premise of the infringement theory is met.
  • A second central question is one of technical scope and equivalence: Does the use of standard, off-the-shelf processor hardware counters (e.g., MPERF/APERF) for general-purpose monitoring and autoscaling in Google Cloud constitute the specific, novel "hardware based utilization metering" apparatus, including an "idle indicator," as claimed in the '809 patent? The case may turn on whether the patent claims a dedicated circuit or a more general hardware-based method.
  • A foundational evidentiary question will be one of plausibility and proof: Given the complaint's significant redactions and reliance on discovery from a prior litigation, a key threshold issue will be whether the publicly available information, combined with the unredacted allegations, is sufficient to plausibly map the complex, proprietary operations of Google Search and Google Cloud to the specific limitations of the asserted claims.
Loading Amended Complaint