by Dennis Crouch
In its decision here, the Federal Circuit took VLSI’s $2.2 billion judgment down to zero — but it will allow a new trial on damages. The court particularly affirmed a finding that Intel infringed claims of U.S. Patent No. 7,523,373, but found the associated $1.5 billion damage award flawed. With regard to the second patent at issue, U.S. Patent No. 7,725,759, the appellate panel found that insufficient particularized evidence had been presented to show infringement under the Doctrine of Equivalents. This took away the $700 million damage for infringement of that second patent. Opinion by Judge Taranto and joined by Judges Lourie and Dyk.
VLSI originally sued Intel in the Western District of Texas in 2019, asserting the two patents relating to semiconductor chip technology. The ‘373 and ‘759 respectively cover “Minimum Memory Operating Voltage Technique” for integrated circuits and a “Method of Managing Clock Speed in an Electronic Device.” And, Intel’s Haswell and Broadwell microprocessors were accused of practicing the claims. A jury issued its verdict in early 2021 favoring the patentee and, the district court entered judgment that Intel literally infringed the ‘373 patent and infringed the ‘759 patent under the doctrine of equivalents with $2.2 billion in total damages.
Infringement under the Doctrine of Equivalents: The most interesting part of the infringement decision focused on the doctrine of equivalents — with the Federal Circuit continuing its efforts to tightly cabin-in the doctrine.
Although the jury sided with the patent holder, the Federal Circuit found VLSI’s evidence legally insufficient to support the verdict. The court explained that “[t]he doctrine of equivalents provides a limited exception to the principle that claim meaning defines the scope of the exclusivity right in our patent system” and is “an ‘exceptional’ basis for liability that must be carefully limited.” The court then highlighted three limiting principles of DoE:
- All Elements Rule: First, proof of equivalents must be limitation specific, not focused only on the claim as a whole, though the limitation-specific inquiry of equivalence may be informed by the “role played by each element in the context of the specific patent claim.”
- Function-Way-Result All Substantially the Same: Second, for the determination of whether a substitute element is only insubstantially different from a claimed element and hence an equivalent, a traditional formulation—appropriate for this case, as VLSI’s use of it indicates—asks “whether a substitute element matches the function, way, and result of the claimed element.” Such matching requires that each of function, way, and result be “substantially the same,” with the “way” requirement of particular importance, as a practical matter, in keeping the doctrine properly limited.
- Particularized Expert Testimony: Third, the patentee must provide “particularized testimony and linking argument as to the insubstantiality of the differences between the claimed invention and the accused device.”
Of course, the court could have also added a fourth and fifth substantial limiter: prosecution history estoppel, and vitiation (which some folks argue is simply a feature of the function-way-result rule). As you read these requirements, you’ll begin to see that the path to winning a case by equivalents is quite narrow — each particular limitation in the claim must be practiced in the accused device, either literally or by equivalents. And, although an equivalent will, by definition, have a different structure than what is claimed, the doctrine expounded by the court here requires that it be “substantially the same” in both its function and structure. And, as noted by the outcome of the case, those substantial similarities are treated as elements of proving equivalents and thus cannot be simply inferred by the jury.
Applying these principles, the court held that VLSI failed to adequately explain why the different way Intel’s processors operated was only an insubstantial difference from what was claimed. During oral arguments, the judges pressed the parties to identify “particularized testimony and linking argument” to support a finding of equivalence — apparently there was not enough. Here, the VLSI claim requires a “first master device” and a distinct “programmable clock controller.” But, in the accused intel products these were overlapping on the same power control unit. VLSI’s expert Dr. Conte testified that “It’s just a difference of where an engineer draws this data line. It’s a design choice.” But, the appellate panel found that this testimony was legally insufficient to provide a basis for the jury to find that spitting functions as Intel did was “substantially the same way” as recited in the claims. Merely labeling it a “design choice” was not enough.
The determination of equivalence under the doctrine of equivalents is a question of fact protected by the 7th Amendment right to a jury verdict. So why was the Federal Circuit reversing the jury’s finding of equivalence in this case? The key guiding principle here is the legal sufficiency of the evidence. Even though infringement under the doctrine of equivalents is a factual issue, the patentee still bears the burden of providing particularized evidence to support its theory of equivalence. As the court stated:
“We have long demanded specificity and completeness of proof as crucial to enforcing the limits on the doctrine: The patentee must provide particularized testimony and linking argument as to the insubstantiality of the differences between the claimed invention and the accused device.”
Essentially, while the ultimate determination is a question for the jury, there is still a minimum legal threshold of providing substantive, thorough evidence explaining why any differences are insubstantial. The Federal Circuit found VLSI failed to meet that threshold with conclusory, unsupported testimony.
This decision continues the court’s strict policing of the doctrine of equivalents — here by overturning a jury verdict and rejecting the district court’s JMOL denial. In a complete reversal from its historic origins, I now see DoE as a highly technical strategy full of almost unavoidable pitfalls.
Damages Award Vacated: Although the Federal Circuit affirmed the literal infringement judgment as to the ‘373 patent, it vacated the $1.5 billion damages award after finding that VLSI’s damages expert made a clear error in calculating the power savings attributable to the patented technology. Specifically, according to the Federal Circuit, the expert used data from processor states that did not practice the patented functionality in order to create a model. Although this issue was presented to the jury, the Federal Circuit found the “readily identifiable” methodology problem required vacating the award. The court remanded for a new trial on damages, allowing VLSI the opportunity to correct this issue. Note here that the patentee strongly disagreed that there was any error, much a “readily identifiable” one sufficient to overturn the jury verdict.
License Defense Permitted
Finally, in a procedural ruling, the Federal Circuit held that the district court improperly denied Intel’s request to amend its answer to add a license defense. Intel alleged that its recent acquisition of Finjan gave it a license to VLSI’s patents under Intel’s earlier agreement with Finjan. The district court found the defense futile and thus refused to permit the amendment, but the Federal Circuit disagreed. While not concluding Intel would ultimately succeed in proving it held a license, the court explained there were open questions under Delaware law that required further litigation. “We hold only that it was error to deny the motion to add the license defense to the case.”