The District of Delaware, Judge Andrews presiding, in Comcast IP Holdings I LLC v. Sprint Communications Co., Civil Action No. 12-205-RGA (D. Del. Sept. 29, 2014), issued a brief order concerning apportionment. Judge Andrews was faced with a motion in limine seeking to exclude profits and and/or revenues relating to the accused products as violating the entire market value rule (EMVR). The court decided that the briefing was not “enough for a good decision, and asked for further submissions including a proffer of the Comcast expert’s testimony.” Slip op. at 1. A proffer was thus requested.

Having received the additional proffer, the court concluded that Comcast’s expert had not properly apportioned. First, the court noted that, had Comcast’s expert started with the accused product revenues and/or profits and apportioned from there, it would have been inclined to allow those numbers into evidence, because “an expert needs a logical starting place for analysis.” Slip op. at 1-2. Second, the court reasoned that, here, Comcast’s expert had not “made the sort of apportionment that the law requires.” Id. at 2. Instead, noted the court, there were only three paragraphs in the proffer that identified three functionalities of the accused products, only one of which implicated the asserted patents. Importantly, the court observed that “[n]on-technology factors contribute to the value of the products,” and that “[u]se patterns contribute to the value.” Id. These factors were apparently on top of the three functionalities, and the court implied that the apportionment would need to account for them. Finally, the court cited the expert’s concession in her deposition that “she did not do a ‘numerical calculation’ to arrive at a percentage to apply to the profits. … Apportionment does not seem possible without a numerical calculation.” The court therefore concluded that Comcast’s expert had failed to comply with the requirement cited from VirnetX, that the party “carefully tie proof of damages to the claimed invention’s footprint in the market place.” Id. (quoting VirnetX).

Two additional points: (1) the court noted that Comcast’s expert had another basis to seek damages in the same amount sought through the provisionally excluded theory; and (2) the court granted Comcast an additional day to submit a proffer with “some detail as to how much of the revenues and/or profits it apportioned to the patented technology, and what the basis for the apportionment is.” Slip op. at 2.

Federal Circuit, in Mentor Graphics Corp. v. Eve-USA, Inc. (also including Synopsys, Inc. as a defendant), Nos. 2015-1470, 2015-1554, 2015-1556 (Fed. Cir. March 16, 2017) (Judge Moore authoring), issued an opinion addressing apportionment and lost profits damages. The case addressed an open question that has gone both ways in district courts: whether lost profits requires further apportionment on top of the Panduit test. In other words, if a patentee is able to prove-up lost profits under Panduit, must the patentee apportion those lost profits to determine the incremental value of the patented feature vis-à-vis any unpatented features? The court answered the question in the negative, but reserved judgment on whether the same outcome would attach in all lost profits cases, particularly where the patentee relied on a theory other than Panduit to prove lost profits. Because the case is replete with meaningful language, the following summary consists mainly of direct quotes.

The patent involved simulation/emulation technology for integrated circuits. More specifically, one of the patents-in-suit, the ‘376 patent, related to debugging source code after synthesis, which is the process of transforming hardware description language (HDL) into gate-level netlists. The patent indicated that prior art simulators only allowed a developer to view the input and output of the netlist and would not allow stepping-through the intermediate gates. Without the latter functionality, it was more difficult to debug at the gate level. The ‘376 patent allowed developers to insert test probes at various stages of a netlist to monitor intermediate values. The court indicated that claim 1 was representative:

  1. A method comprising the steps of:

a) identifying at least one statement within a register transfer level (RTL) synthesizable

source code; and

b) synthesizing the source code into a gate-level

netlist including at least one instrumentation

signal, wherein the instrumentation signal is

indicative of an execution status of the at least

one statement.

After affirming the lower court’s denial of Synopsys’ JMOL motion of noninfringement, the court turned to lost profits damages. (Between those sections is a brief analysis of assignor estoppel.) “At trial, Mentor argued it was entitled to obtain lost profit damages for lost sales of its Veloce emulators resulting from Synopsys’ infringing sales of its ZeBu emulators because Mentor would have made additional Veloce sales but for Synopsys’ infringing ZeBu sales.” Slip op. at 8. The district court instructed the jury about lost profits, including “extensive discussion of each of the Panduit factors.” Id. The jury awarded over $36M in lost profits damages. Synopsys appealed arguing that “the damage award should be vacated because the district court failed to apportion the lost profits.” Id. The Federal Circuit did not agree.

Key quotes regarding lost profits follow:

Compensatory damages are a staple across most every area of law. And compensatory damages under the patent statute, which calls for damages adequate to compensate the plaintiff for its loss due to the defendant’s infringement, should be treated no differently than the compensatory damages in other fields of law. … When a plaintiff proves it would have been in a certain position but for a defendant’s harmful act, it is entitled to damages to put it in the same position it would have occupied had the harmful act never occurred. … The “but for” damages the patentee must establish in patent law, as the Supreme Court explained, are an answer to a simply stated question: “[H]ad the Infringer not infringed, what would the Patent Holder-Licensee have made?” Aro Mfg. Co., 377 U.S. at 507.

There is no particular required method to prove but for causation. One “useful, but non-exclusive” method to establish the patentee’s entitlement to lost profits is the Panduit test first articulated by the Sixth Circuit. … The goal of lost profit damages is to place the patentee in the same position it would have occupied had there been no infringement. … Thus, the fact finder’s job is to determine what would the patent holder have made (what would his profits have been) if the infringer had not infringed.

Under the Panduit test, a patentee is entitled to lost profit damages if it can establish four things:

(1) demand for the patented product;

(2) absence of acceptable non-infringing alternatives;

(3) manufacturing and marketing capability to exploit

the demand; and

(4) the amount of profit it would have made.

… Damages under Panduit are not easy to prove. …

We have explained the relationship between the first two Panduit factors. The first factor—demand for the patented product—considers demand for the product as a whole. The second factor—the absence of non-infringing alternatives—considers demand for particular limitations or features of the claimed invention. Together, requiring patentees to prove demand for the product as a whole and the absence of non-infringing alternatives ties lost profit damages to specific claim limitations and ensures that damages are commensurate with the value of the patented features.

The second factor, absence of acceptable non-infringing alternatives, often proves the most difficult obstacle for patent holders. Under this factor, if there is a noninfringing alternative which any given purchaser would have found acceptable and bought, then the patentee cannot obtain lost profits for that particular sale. For example, if the customer would have bought the infringing product without the patented feature or with a different, non-infringing alternative to the patented feature, then the patentee cannot establish entitlement to lost profits for that particular sale. And this determination is made on a customer-by-customer basis. For this reason, it is quite common to see damage awards where, as in this case, the patentee proves entitlement to lost profits for some of its sales, but not others. For sales in which the patentee cannot prove the elements necessary to establish entitlement to lost profits, the statute guarantees the patentee a reasonable royalty for those sales. …

Slip op. at 10-15 (emphasis added) (citations and footnotes omitted).

The court went on to address the specific facts of the case at hand. One point that should be emphasized: “In this case, apportionment was properly incorporated into the lost profits analysis and in particular through the Panduit factors. Panduit’s requirement that patentees prove demand for the product as a whole and the absence of non-infringing alternatives ties lost profit damages to specific claim limitations and ensures that damages are commensurate with the value of the patented features. We leave for another day whether a different theory of ‘but for’ damages adequately incorporates apportionment principles. We hold today that on the undisputed facts of this record, satisfaction of the Panduit factors satisfies principles of apportionment: Mentor’s damages are tied to the worth of its patented features.” Slip op. at 17-18.

Some additional quotes from the court:

The facts of this case are remarkably simple for a patent damages appeal and Synopsys does not dispute any of them. The relevant market (suppliers of emulators to Intel) contained two parties, Synopsys and Mentor. Mentor sold its own Veloce emulators to Intel and Synopsys sold its ZeBu emulators to Intel which were found to infringe Mentor’s ’376 patent claims. Synopsys does not dispute that but for its infringement, Mentor would have made each of the infringing emulator sales to Intel. Nor does it dispute how much Mentor would have earned, the precise numbers of sales Mentor would have made, whether there were any alternatives that Intel may have preferred over the purchase of Mentor’s product, or whether Intel would have chosen to purchase fewer emulators. In short, Synopsys does not dispute on appeal that for each infringing sale it made to Intel, Mentor lost that exact sale.

This is important as it makes this case quite narrow and unlike the complicated fact patterns that impact so many damages models in patent cases. The jury found, and Synopsys does not dispute on appeal, that Mentor satisfied all of the Panduit factors with regard to the sales to Intel for which the jury awarded lost profits … Synopsys does not challenge the sufficiency of Mentor’s evidence with regard to the individual Panduit factors. In this case, the jury found, and Synopsys does not dispute, that Intel would not have purchased the Synopsys emulator system without the two patented features and that there were no other alternatives available. Despite hearing evidence that there were many valuable and important features in the emulator system, this jury found that if Synopsys could not have sold its emulator system with the two infringing features (Mentor’s patented features), Intel would have bought the emulators from Mentor. There were no other competitors, and the jury found there were no non-infringing alternative emulator systems which would have satisfied Intel. Thus, what did Mentor lose when Synopsys appropriated its two patented features? It lost the profits it would have made on the sale of its emulators to Intel. These are the simple, undisputed facts on appeal.

Synopsys largely ignores these facts and seeks to have us depart from basic compensatory damages principles equally applied across many areas of law. Synopsys advocates for a two-step process for calculating lost profits. First, Synopsys argues a patentee must calculate the amount of profits it lost as a result of the infringement using the Panduit factors. Second, Synopsys argues a patentee must further apportion its lost profits to cover only the patentee’s inventive contribution. … Synopsys argues that the allegedly infringing features were just two features of emulators that comprise thousands of hardware and software features. Thus, according to Synopsys, Mentor is not entitled to recover what it lost, the amount necessary to make it whole for the sales it lost, but rather the value attributable to its patented features.

Synopsys argues that “[p]rinciples of apportionment play an especially vital role in this age of complex, multicomponent electronic devices.” Synopsys

argues that the patentee does not “deserve,” lost profits for the whole emulator when it only invented some of the features on the emulator. Thus, according to Synopsys the damages should not be the profits the patentee lost when it lost the emulator sale because of Synopsys’ infringement, but rather only the amount of profit properly attributable to its patented features.

We agree with Synopsys that apportionment is an important component of damages law generally, and we believe it is necessary in both reasonable royalty and lost profits analysis. In this case, apportionment was properly incorporated into the lost profits analysis and in particular through the Panduit factors. Panduit’s requirement that patentees prove demand for the product as a whole and the absence of non-infringing alternatives ties lost profit damages to specific claim limitations and ensures that damages are commensurate with the value of the patented features. We leave for another day whether a different theory of “but for” damages adequately incorporates apportionment principles. We hold today that on the undisputed facts of this record, satisfaction of the Panduit factors satisfies principles of apportionment: Mentor’s damages are tied to the worth of its patented features.

The jury found, and Synopsys does not dispute, there were only two acceptable alternatives to Intel: Mentor’s emulator and Synopsys’ infringing emulator. The jury was properly instructed that if there were any other acceptable, non-infringing emulation system or if there were prototypes that may have been acceptable or if there was any acceptable non-infringing alternative that could have been made available (even if they did not already exist), then Mentor could not receive lost profits on those particular sales. J.A. 164.8 The jury was also instructed that it could have found the patented features were not critical to Intel and that it would have purchased Synopsys’ emulators without the features. The instruction expressly stated that Mentor could not obtain lost profit damages if “Synopsys could have made available during the damages period an acceptable, non-infringing alternative to Mentor-Graphics’ emulation system and Synopsys’ infringing emulation system.” J.A. 164. Synopsys could have made its emulator system minus the two infringing features—that would have been an alternative to the “Synopsys infringing emulation system.” However, the jury concluded, and Synopsys does not dispute on appeal the jury’s Panduit fact findings, that there was no such non-infringing alternative that Intel would have purchased.


Synopsys and the amicus brief argue that complex multi-feature devices necessitate change in patent damages law. They argue that not requiring an additional apportionment step after the Panduit test has been met would “allow multiple entities to obtain lost profits on the same product where each entity holds a patent on a different ‘but for’ feature of the same product.” Amicus Br. 11. This claimed threat of “serial infringement claims” is not correct. Again, we do not speak to all damages models. Under Panduit, however, there can only be one recovery of lost profits for any particular sale.

The jury found (and Synopsys does not challenge on appeal) that Intel would not have purchased emulators without the features claimed in Mentor’s ’376 patent. While there may have been other features of the emulator that were important to Intel, only Mentor could sell Intel an emulator with all the features it required. Because Mentor had proprietary rights to the only means of satisfying this demand by Intel, because no other party could sell Intel an emulator with those two components, no one else had the right to sell emulators to Intel that satisfied all of Intel’s requirements. In short, for these particular sales, no other party could satisfy the Panduit factors, making it impossible for multiple patentees to obtain lost profit damages for the same sales.


With such multi-component products, it may often be the case that no one patentee can obtain lost profits on the overall product—the Panduit test is a demanding one. A patentee cannot obtain lost profits unless it and only it could have made the sale—there are no non-infringing alternatives or, put differently, the customer would not have purchased the product without the infringing feature.

Consider the laptop example. If the only patented component is the extended life battery and a customer will only buy a laptop with this battery (meaning a laptop with a lower quality battery is not an acceptable noninfringing alternative to the customer), then when an infringer who appropriates the patented extended life battery sells a laptop, the infringer has deprived the patentee of the lost profits on the laptop sale which only it could have made. If a laptop with a lower-quality battery would be an acceptable non-infringing alternative to certain customers, the patentee would not be entitled to lost profits for these laptop sales. For those customers, the patented battery was not a factor in their purchasing decision; it was not necessary for the sale. The only sales for which the patentee can obtain lost profits are the customers who would refuse to purchase laptops without the patented extended-life battery. For these lost customers, the extended-life battery drives their purchasing decisions.

We hold that the district court did not err in refusing to further apportion lost profits after the jury returned its verdict applying the Panduit factors. We conclude that, when the Panduit factors are met, they incorporate into their very analysis the value properly attributed to the patented feature. We affirm the district court’s denial of judgment as a matter of law and/or motion for new trial with regard to damages.

Slip op. at 15-22 (emphasis added) (citations and footnotes omitted).