DCT
3:18-cv-00326
Tabaian v. Intel Corp
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Fereydun Tabaian and Ahmad Ashrafzadeh (California)
- Defendant: Intel Corporation (Delaware)
- Plaintiff’s Counsel: Klarquist Sparkman LLP; Wright Close & Barger, LLP; Laminack, Pirtle & Martines; Law Offices of Luke Motley, IV PC
- Case Identification: 3:18-cv-00326, D. Or., 02/21/2018
- Venue Allegations: Venue is alleged to be proper in the District of Oregon because Intel has committed acts of infringement in the district and maintains numerous regular and established places of business there.
- Core Dispute: Plaintiffs allege that Defendant’s microprocessors incorporating Fully Integrated Voltage Regulator (FIVR) technology infringe a patent related to the programmable calibration of on-chip power supplies.
- Technical Context: The technology concerns on-die voltage regulation for microprocessors, a method for improving power efficiency and performance by precisely managing power delivery directly on the processor chip itself.
- Key Procedural History: The complaint alleges that Plaintiffs disclosed their invention to Intel under a Non-Disclosure Agreement in 2003, several years before the patent-in-suit issued. The complaint further alleges that after Intel declined to pursue the technology, it proceeded to develop its accused FIVR technology and that Intel engineers later cited the patent-in-suit in their own patent applications, a fact pattern that may be used to support allegations of pre-suit knowledge and willful infringement.
Case Timeline
| Date | Event |
|---|---|
| 2003-06-30 | Provisional patent application filed for the '944 invention |
| 2003-07-11 | Plaintiffs and Intel enter into a Non-Disclosure Agreement |
| 2003-07 to 09 | Plaintiffs present details of the '944 invention to Intel |
| 2006-04-11 | U.S. Patent No. 7,027,944 issues |
| 2013-06-01 | Alleged start of infringing conduct with launch of first FIVR products |
| 2018-02-21 | Complaint for Patent Infringement filed |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 7,027,944 - Programmable Calibration Circuit and Power Supply Current Sensing and Droop Loss Compensation, issued April 11, 2006
The Invention Explained
- Problem Addressed: The patent’s background section describes challenges in multi-phase voltage regulators used for microprocessors, including unequal current sharing between phases, inaccuracies in voltage "droop" (an intentional voltage reduction under load), and performance variations caused by temperature changes and manufacturing inconsistencies ('944 Patent, col. 1:12-44). These issues can lead to inefficient power use and necessitate the discarding of processors that do not meet power specifications, which is described as "very inefficient and costly" ('944 Patent, col. 1:63-65).
- The Patented Solution: The invention is a programmable calibration circuit that works with a voltage regulator to correct for these inaccuracies. As detailed in the specification and illustrated in Figure 2, the circuit includes a controller (500) that receives inputs regarding the regulator's load voltage (570) and operating temperature (580). Using calibration data stored in nonvolatile memory (590), the controller adjusts the regulator's behavior—specifically its "sense outputs" and "droop outputs"—to ensure precise and consistent power delivery tailored to the specific processor it is powering ('944 Patent, Abstract; col. 2:11-29).
- Technical Importance: This approach allows for the dynamic, post-manufacturing calibration of a processor's power supply, enabling a regulator to be precisely tuned to a specific chip’s unique power requirements and to compensate for performance drift over temperature ('944 Patent, col. 5:1-9).
Key Claims at a Glance
- The complaint asserts infringement of independent claim 1 and dependent claims 2, 5-11, 13-14, 16, and 18-25 (Compl. ¶33).
- Independent Claim 1 recites a circuit with the following essential elements:
- A regulator circuit and a calibration control circuit.
- The calibration control circuit includes a controller, an interface with nonvolatile memory, droop outputs, sense outputs, a load voltage input, and a temperature input.
- The nonvolatile memory stores calibration data.
- The calibration control circuit interfaces with the regulator circuit (via sense, droop, and load voltage connections), with the nonvolatile memory (to store data), and with the temperature and load voltage inputs (to receive data and to calibrate the stored data).
- The received temperature data is used by the calibration control circuit to adjust the sense outputs and droop outputs.
III. The Accused Instrumentality
Product Identification
The accused instrumentalities are Intel processors that incorporate its "Fully Integrated Voltage Regulator ('FIVR')" technology (Compl. ¶4). The complaint specifically names the Haswell, Broadwell, Ice Lake, Tigerlake, and certain Skylake processor families (Compl. ¶4).
Functionality and Market Context
The complaint alleges that the accused FIVR is a high-frequency, multi-phase voltage regulator integrated directly onto the processor die (Compl. ¶34). The infringing functionality is alleged to reside in the "Power Control Unit ('PCU')" and/or "FIVR Control Module ('FCM')", which the complaint identifies as the "calibration control circuit" (Compl. ¶38). This PCU/FCM is alleged to use temperature sensor data to determine an "offset VID [Voltage ID] value," which is then used to calculate a calibrated voltage setting for the regulator (Compl. ¶45). This process allegedly relies on calibration data, such as optimal voltage-frequency curves and load line parameters, stored in nonvolatile memory (Compl. ¶¶49, 51, 61). The complaint positions FIVR as a key enabling technology for significant market advantages, including increased battery life, higher peak performance, and manufacturing cost reductions (Compl. ¶¶25-27).
IV. Analysis of Infringement Allegations
No probative visual evidence provided in complaint.
'944 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| A circuit comprising: a regulator circuit and a calibration control circuit... | The accused Intel processors contain a FIVR, which is alleged to be a regulator circuit, and a Power Control Unit (PCU) and/or FIVR Control Module (FCM), which are alleged to comprise a calibration control circuit (Compl. ¶¶37-38). | ¶37, ¶38 | col. 2:11-13 |
| wherein said calibration control circuit includes a controller, an interface with nonvolatile memory, droop outputs, sense outputs, load voltage input, and temperature input | The PCU/FCM is alleged to include a controller and to interface with nonvolatile memory, utilize droop and sense outputs, and receive load voltage and temperature inputs to manage the FIVR (Compl. ¶¶39, 41, 52). | ¶39, ¶41, ¶52 | col. 2:13-16 |
| wherein... said nonvolatile memory stores calibration data | Nonvolatile memory in the accused products is alleged to store calibration data, including VID offset values as a function of temperature and current, and optimal voltage-frequency curves for processor cores (Compl. ¶¶49, 51, 53). | ¶49, ¶51, ¶53 | col. 2:22-24 |
| wherein... said temperature data is used by said calibration control circuit to adjust said sense outputs and said droop outputs | The complaint alleges that Intel implements calibration by determining an offset VID value based on temperature sensor data, which adjusts the output voltage and implements the droop function, thereby adjusting the sense and droop outputs (Compl. ¶¶45, 48, 57). | ¶45, ¶48, ¶57 | col. 2:19-22 |
| wherein... said calibration control circuit interfaces with said temperature input and said load voltage input to calibrate said calibration data stored in said nonvolatile memory | The PCU is alleged to interface with temperature and load voltage inputs to perform a calibration process that determines optimal voltage-frequency curves and VID offset values, which are then stored as calibration data in nonvolatile memory (Compl. ¶¶49, 51, 58). | ¶49, ¶51, ¶58 | col. 2:24-27 |
Identified Points of Contention
- Scope Questions: A central dispute may be whether Intel's integrated Power Control Unit (PCU), a complex, multi-functional component of the processor, can be properly characterized as the "calibration control circuit" recited in the patent. The defense may argue the patented circuit is a more specialized, standalone component as depicted in the patent's figures, whereas the PCU performs many functions beyond calibration.
- Technical Questions: The complaint alleges that using temperature data to calculate an "offset VID value" (Compl. ¶45) satisfies the claim limitation "temperature data is used by said calibration control circuit to adjust said sense outputs and said droop outputs." A key technical question will be whether this indirect adjustment via a VID code is the same as the more direct adjustment of amplifier circuits described in the patent's preferred embodiment ('944 Patent, FIG. 1, items 150, 180).
V. Key Claim Terms for Construction
"calibration control circuit"
- Context and Importance: This term defines the core of the invention. Its construction will be critical, as infringement hinges on whether Intel's PCU/FCM falls within its scope. Practitioners may focus on this term because the complaint equates it with Intel's highly integrated PCU, which may have a different architecture and broader functionality than the circuit described in the patent's embodiments.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The claim itself defines the circuit functionally by listing its required inputs, outputs, and interfaces (Claim 1). Plaintiffs may argue that any circuit performing these functions, regardless of its integration level or other capabilities, meets the definition.
- Evidence for a Narrower Interpretation: The patent specification describes a specific implementation in FIG. 2, showing a controller (500) with distinct Digital-to-Analog and Analog-to-Digital converters for its various interfaces. A defendant could argue that the term should be limited to a circuit with a similar, dedicated architecture, rather than a general-purpose processor control unit.
"adjust said sense outputs and said droop outputs"
- Context and Importance: This phrase describes the mechanism of action for the calibration. The dispute will likely focus on what constitutes an "adjustment." The complaint alleges an indirect mechanism (calculating an offset VID), and the viability of this theory depends on the construction of "adjust."
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: Plaintiffs may argue that any action that ultimately causes the sense and droop functions of the regulator to change in response to temperature data constitutes an "adjustment," including modifying a central VID code that governs the regulator's overall behavior.
- Evidence for a Narrower Interpretation: A defendant may point to the patent's description, where the controller's outputs appear to directly drive and modify specific "adjustable sense amplifier[s]" and an "adjustable droop amplifier" ('944 Patent, FIG. 1). This could support a narrower construction requiring a more direct control mechanism than what is alleged against the accused products.
VI. Other Allegations
Indirect Infringement
The complaint includes a general allegation of infringement under 35 U.S.C. § 271(b) (inducement), suggesting that by selling the accused processors to computer manufacturers, Intel induces them to make and sell infringing systems (Compl. ¶29).
Willful Infringement
The complaint lays a detailed foundation for willfulness. It alleges that Intel had pre-suit knowledge of the invention itself through a series of meetings with the inventors under a Non-Disclosure Agreement in 2003 (Compl. ¶¶19-22). It further alleges that after the '944 patent issued, Intel engineers cited it as prior art in at least eight of their own patent applications, suggesting knowledge of the patent itself (Compl. ¶23).
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of claim scope: can the term "calibration control circuit," as defined in the patent, be construed to read on a complex, multi-functional, and deeply integrated processor component like Intel's Power Control Unit, or is it limited to a more dedicated circuit architecture as shown in the patent's embodiments?
- A key evidentiary question will be one of technical mechanism: does Intel's alleged method of using temperature data to calculate an "offset VID value" that governs the regulator's output constitute the specific claimed function of using temperature data to "adjust said sense outputs and said droop outputs," or is this a fundamentally different and non-infringing process?
- Finally, the case will likely feature a significant focus on willfulness and pre-suit conduct: should infringement be found, a central question for the fact-finder will be whether the alleged pre-suit history—including disclosure under an NDA and subsequent development of the accused technology—is sufficient to establish that any infringement by Intel was willful.