DCT

3:17-cv-00187

University Of Tennessee Research Foundation v. SAP America Inc

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 3:17-cv-00187, E.D. Tenn., 05/02/2017
  • Venue Allegations: Plaintiff alleges venue is proper because Defendant SAP is registered to do business in Tennessee, has transacted business within the Eastern District of Tennessee, and has allegedly committed acts of infringement in the district.
  • Core Dispute: Plaintiff alleges that Defendant’s SAP HANA and SAP Adaptive Server Enterprise database products infringe five patents related to parallel data processing, workload balancing, and data clustering for large-scale databases.
  • Technical Context: The patents address foundational methods for improving database performance by distributing computational loads across multiple processors and organizing data based on its structural properties, technologies central to the "big data" and high-performance computing sectors.
  • Key Procedural History: The complaint alleges that the patents-in-suit have been cited by over 300 U.S. patents and applications from numerous major technology companies, which Plaintiff presents as evidence of the inventions' importance. It also alleges pre-suit knowledge for U.S. Patent Nos. 6,741,983 and 7,272,612 based on citations in patents assigned to Defendant's parent company, SAP SE.

Case Timeline

Date Event
1999-09-28 Priority Date for all five patents-in-suit
2004-05-25 U.S. Patent No. 6,741,983 Issues
2007-09-18 U.S. Patent No. 7,272,612 Issues
2008-11-18 U.S. Patent No. 7,454,411 Issues
2011-02-01 U.S. Patent No. 7,882,106 Issues
2012-01-17 U.S. Patent No. 8,099,733 Issues
2017-05-02 Complaint Filed

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

U.S. Patent No. 7,454,411 - "Parallel Data Processing Architecture"

The Invention Explained

  • Problem Addressed: The patent family addresses the problem that standard database indexing structures fail to provide “rapid access to records” or “take advantage of naturally occurring structure in the data” when dealing with vast, multi-dimensional datasets, such as DNA profiles (Compl. ¶5; ’983 Patent, col. 1:48-51).
  • The Patented Solution: The invention describes a parallel computer architecture for databases designed to distribute and balance computational workloads. It employs a system with a "root host processor" to handle client queries and a plurality of other "host processors." These processors communicate their current load and capacity to one another and exchange search requests to keep their individual work queues balanced, thereby minimizing the time required to perform a search (Compl. ¶¶ 32, 63, 65; ’411 Patent, Abstract; ’411 Patent, col. 27:6-14).
  • Technical Importance: This architectural approach enables database systems to scale efficiently and handle the exponentially growing volumes of data characteristic of modern applications (Compl. ¶¶ 3, 11).

Key Claims at a Glance

  • The complaint asserts at least independent claim 1 (Compl. ¶71).
  • The essential elements of claim 1 include:
    • A parallel data processing system with a plurality of host processors, including a root host processor responsive to client queries.
    • At least two host processors having a queue of search requests, each executing a search engine.
    • The host processors communicate capacity and load information and exchange at least one search request.
    • The system includes a communications system coupling the host and root processors.
    • Selected host processors store a database index.
  • The complaint does not explicitly reserve the right to assert dependent claims.

U.S. Patent No. 8,099,733 - "Parallel Data Processing Architecture"

The Invention Explained

  • Problem Addressed: As with the related ’411 Patent, this invention addresses the challenge of distributing computational loads evenly across multiple nodes in a parallel computing platform to avoid performance bottlenecks when querying large databases (Compl. ¶¶ 3, 10).
  • The Patented Solution: The patent teaches a method for balancing workloads where host processors broadcast their current load (e.g., search queue length) and capacity to other processors. Recipient processors then use this information to bring their search queues "into balance with another...according to a time constant" (’733 Patent, col. 32:58-63). This is achieved by exchanging blocks of search requests from hosts with longer wait times to hosts with shorter wait times (Compl. ¶¶ 36-39).
  • Technical Importance: The claimed method aims to equalize the average wait time for computation of requests across all processors, improving overall system responsiveness and throughput (Compl. ¶39).

Key Claims at a Glance

  • The complaint asserts at least independent claim 18 (Compl. ¶98).
  • The essential elements of claim 18 include:
    • A method of balancing workload between a plurality of host processors.
    • Maintaining load information of processor capacity and search queue length.
    • Broadcasting this load information to at least one other host processor.
    • Bringing a search queue of client queries into balance with another host processor according to a time constant responsive to the broadcast information.
  • The complaint does not explicitly reserve the right to assert dependent claims.

U.S. Patent No. 6,741,983 - "Method of Indexed Storage and Retrieval of Multidimensional Information"

  • Technology Synopsis: The patent teaches a method for organizing data records into clusters to improve search efficiency. It involves representing data variables in binary form and using principal component analysis to identify vectors that yield distinct data clusters, allowing a test to be formulated that assigns each data record to a specific cluster (Compl. ¶¶ 43-45, 110-115; ’983 Patent, Abstract).
  • Asserted Claims: At least independent claim 1 (Compl. ¶119).
  • Accused Features: The complaint alleges that SAP HANA products infringe by using technology to represent variables in binary form, apply principal component analysis, and formulate tests based on selected vectors to organize data records into clusters (Compl. ¶¶ 110-115).

U.S. Patent No. 7,272,612 - "Method of Partitioning Data Records"

  • Technology Synopsis: The patent describes a method for partitioning data records into groups by defining a function based on a combination of "entropy" and "adjacency" measures for a designated variable. An optimization procedure is applied to this function to determine how to partition the values of the variable, and thereby the data records, into two or more groups (Compl. ¶¶ 48-49, 131-132; ’612 Patent, Abstract).
  • Asserted Claims: At least independent claim 1 (Compl. ¶137).
  • Accused Features: The complaint alleges that SAP HANA infringes by partitioning data across multiple hosts using a function that combines measures of entropy and adjacency, which is then optimized to assign data records to groups (Compl. ¶¶ 131-133).

U.S. Patent No. 7,882,106 - "Method of Indexed Storage and Retrieval of Multidimensional Information"

  • Technology Synopsis: This patent is related to the ’612 Patent and similarly teaches a method for partitioning data records in a multi-dimensional database. The method uses a function comprising a combination of entropy and adjacency measures, with the adjacency measure weighted by a factor, to partition records into groups for storage in computer memory (Compl. ¶¶ 53-54, 150-151; ’106 Patent, Abstract).
  • Asserted Claims: At least independent claim 14 (Compl. ¶155).
  • Accused Features: The complaint accuses SAP HANA's parallel data processing system of infringing by partitioning data records of multi-dimensional databases using a function that combines weighted measures of entropy and adjacency (Compl. ¶¶ 149-151).

III. The Accused Instrumentality

Product Identification

  • SAP HANA 1.0 SPS05-SPS12 and SAP HANA 2.0 SPS00 (accused of infringing the ’411, ’983, ’612, and ’106 patents) (Compl. ¶¶ 58, 107, 129, 147).
  • SAP Adaptive Server Enterprise (ASE) (versions 15.7, 15.5, 15.0.1, and 15.0.3 Cluster Edition) (accused of infringing the ’733 patent) (Compl. ¶80).

Functionality and Market Context

  • The complaint describes SAP HANA as a system for "high-performing analysis on huge volumes of granular data" using parallel processing (Compl. ¶60, p. 20 visual). A provided diagram shows that its parallel query processing involves splitting an execution plan into multiple operators that are executed in parallel and can be further broken down into jobs dispatched to multiple threads. This diagram illustrates a multi-user, multi-query parallel processing system (Compl. p. 21 visual).
  • The complaint describes SAP ASE Cluster Edition as a system where "multiple Adaptive Servers...run as a shared-disk cluster" connected by a "high-speed private interconnection" (Compl. p. 28 visual). A diagram depicts a network topology for ASE with four nodes connected via "Primary Private" and "Secondary Private" networks, over which the complaint alleges they broadcast load information (Compl. p. 26 visual).
  • Plaintiff alleges these products are part of a market trend requiring new technologies to analyze "big data sets at terabyte or even petabyte scale" (Compl. ¶11).

IV. Analysis of Infringement Allegations

'411 Patent Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
a parallel data processing system for search, storage, and retrieval of data of a database...comprising a plurality of host processors including a root host processor... SAP HANA is alleged to be a parallel processing system that natively performs analysis on data from SAP and non-SAP systems, comprising a plurality of host processors including a root host processor (Compl. p. 20 visual). ¶¶60, 61 col. 30:2-6
wherein at least two host processors of said plurality of host processors having a queue of search requests...each executing a search engine... The SAP '411 Products are alleged to include at least two host processors with a search engine and a search queue for client queries (Compl. p. 21 visual). ¶62 col. 30:10-12
communicating capacity and load information between said at least two host processors...and said at least two host processors exchanging at least one search request... The SAP '411 Products allegedly include at least two host processors that communicate capacity and load information and exchange at least one search request (Compl. p. 21 visual). ¶63 col. 30:13-17
a communications system coupling said host processors and said root processor... The SAP '411 Products allegedly include a communications system coupling the host and root processors (Compl. p. 20 visual). ¶66 col. 30:18-20
selected host processors of said plurality of host processors storing a database index for said database... The SAP '411 Products are alleged to allow selected host processors to store a database index comprising nodes of a database tree (Compl. p. 20 visual). ¶67 col. 30:21-23
  • Identified Points of Contention:
    • Scope Questions: A potential issue is whether SAP HANA's architecture, as depicted in the complaint's diagrams (Compl. pp. 20-21), has a designated "root host processor" that is structurally and functionally distinct from other "host processors" as required by the claim. The defense may argue that the system operates in a more distributed or peer-to-peer fashion that does not map onto the claim's specific root/host hierarchy.
    • Technical Questions: The complaint alleges that host processors "maintain information of a search queue" and "exchang[e] at least one search request" (Compl. ¶¶ 62-63). A key question will be what evidence supports these specific functions, as the provided diagrams illustrate high-level query plan parallelization rather than the explicit management and exchange of requests between distinct processor queues as claimed.

'733 Patent Infringement Allegations

Claim Element (from Independent Claim 18) Alleged Infringing Functionality Complaint Citation Patent Citation
A method of balancing workload between said plurality of host processors responsive to a client query... The SAP '733 Products are alleged to comprise a method for balancing workload between a plurality of host processors (Compl. p. 26 visual). ¶85 col. 32:51-53
maintaining load information of processor capacity and search queue length for each of said plurality of host processors... The SAP '733 Products are alleged to maintain load information of processor capacity and search queue length for each host processor (Compl. p. 26 visual). ¶86, 88 col. 32:54-56
broadcasting said load information of processor capacity and search queue length of at least one of said plurality of host processors... SAP ASE Cluster Edition is alleged to broadcast load information over "Primary Private" and "Secondary Private" networks connecting the nodes (Compl. p. 26 visual). ¶84, 89 col. 32:57-59
bringing its search queue of client queries into balance with another of said plurality of host processors according to a time constant responsive to receipt... The SAP '733 Products are alleged to enable each host processor to bring its search queue into balance with another host processor according to a time constant responsive to the receipt of the broadcast capacity and load information (Compl. p. 28 visual). ¶91 col. 32:60-63
  • Identified Points of Contention:
    • Scope Questions: A central question will be the construction of "according to a time constant." The defendant may argue that its load balancing mechanism in ASE, while responsive to load information, does not use a specific "time constant" as taught in the patent's specification, but rather employs a different algorithm.
    • Technical Questions: The complaint alleges that ASE balances search queues by "exchanging unprocessed search requests with a recipient host processor responsive to a stochastic selection process" (Compl. ¶92). The provided architectural diagrams (Compl. pp. 26-28) do not detail the specific algorithm used for workload distribution. A key evidentiary question will be whether the complaint provides sufficient factual support for this specific "stochastic" process.

V. Key Claim Terms for Construction

  • Term ('411 Patent, Claim 1): "root host processor"

    • Context and Importance: This term appears to define a specific role within the claimed architecture, distinct from other "host processors." The infringement analysis may turn on whether the accused SAP HANA system has a component that meets this structural and functional definition, or if its query processing is distributed in a way that lacks such a designated "root."
    • Intrinsic Evidence for Interpretation:
      • Evidence for a Broader Interpretation: The specification states the "root host processor...is responsive to client queries" (’411 Patent, col. 3:21-22). This could be argued to cover any processor that initially receives a query before it is parallelized.
      • Evidence for a Narrower Interpretation: The specification also describes the root host processor creating a "search client object" and establishing an "initial search queue" (’411 Patent, col. 3:22-25), suggesting a specific, dedicated role in initiating and managing the query process beyond simply being the entry point.

  • Term ('733 Patent, Claim 18): "balancing...according to a time constant"

    • Context and Importance: This phrase suggests a specific type of control mechanism for achieving workload balance, rather than any generic balancing method. Practitioners may focus on this term because infringement will depend on whether the accused ASE product uses a balancing algorithm that can be properly characterized as being governed by a "time constant" responsive to broadcast load information.
    • Intrinsic Evidence for Interpretation:
      • Evidence for a Broader Interpretation: The term is not explicitly defined, which may support an argument that it should be given its plain and ordinary meaning in the context of control systems, potentially covering various feedback loop mechanisms.
      • Evidence for a Narrower Interpretation: The specification states that a "time constant is utilized to avoid excessive oscillations in host loading" (’733 Patent, col. 28:11-12). This context suggests the term refers to a specific parameter in a dynamic control algorithm designed to manage the rate of response to load imbalances, potentially narrowing its scope to exclude systems that use different stability mechanisms.

VI. Other Allegations

  • Indirect Infringement: The complaint alleges inducement of infringement for all five patents. The allegations are based on Defendant providing the accused products with the capability to operate in an infringing manner, along with "documentation and training materials," "user manuals, product support, [and] marketing materials" that allegedly instruct customers and end-users to utilize the products in a way that directly infringes (Compl. ¶¶ 74, 101, 123, 141, 158).
  • Willful Infringement: The complaint alleges willful infringement for all patents. For the ’411, ’733, and ’106 patents, the allegation is based on knowledge acquired "at least service of this Complaint or shortly thereafter" (Compl. ¶¶ 73, 100, 157). For the ’983 and ’612 patents, the complaint alleges pre-suit knowledge based on patents assigned to Defendant's parent company, SAP SE, that cite the asserted patents as prior art, with the earliest alleged knowledge date being January 24, 2012 (Compl. ¶¶ 121, 139).

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

  • A core issue will be one of architectural mapping: do the high-level marketing and technical diagrams provided for SAP HANA and SAP ASE accurately reflect an architecture that meets the specific structural limitations of the asserted claims, such as the "root host processor" hierarchy of the '411 patent, or is there a fundamental mismatch?
  • A key evidentiary question will be one of functional specificity: does the complaint provide sufficient factual basis to show that the accused products' load-balancing functions perform the specific algorithms required by the claims—such as balancing "according to a time constant" or using a "stochastic selection process"—or do they use functionally distinct methods to achieve a similar outcome?
  • A central question for willfulness will be whether prior art citations in patents assigned to a defendant's parent company can establish pre-suit knowledge of infringement for the '983 and '612 patents, an issue that will depend on evidence of information flow and awareness within the corporate structure.