Patent Damages Apportionment and the Cornell Case

In the high stakes battle of patent litigation, one of the hottest topics in damages is patent value apportionment - the concept that a patentee can only claim the value of the patent against an infringer.  While damages must be limited to the patented feature, the courts also recognize the Entire Market Value Rule (EMVR) exception: damages can be awarded from both patented and unpatented elements of a product only when the patented feature is the basis for demand for the entire product.

The last three Congresses have seen similar Patent Reform Bills, with the current bill calling for a calculation of reasonable royalty damages via one of three methods:  1) Via established marketplace licensing similar to the use made of the invention by the infringer; 2) “Upon a showing to the satisfaction of the court that the claimed invention’s specific contribution over the prior art is the predominant basis for market demand for an infringing product or process, damages may be based upon the entire market value of that infringing product or process;” or 3) If neither of the previous methods are used, the court “shall conduct an analysis to ensure that a reasonable royalty is applied only to the portion of the economic value of the infringing product or process properly attributable to the claimed invention’s specific contribution over the prior art.” S.515, 2009, Sec 4(c)(1).

While none of the bills have come to a vote, courts seem to have received the message.  Courts are increasingly spending more time analyzing the apportionment issue, and one often discussed case is Cornell University v. Hewlett-Packard Company, 609 F. Supp. 2d 279 (N.D.N.Y. 2009). This case was decided at the district level in March 2009, but it has the distinction of having been decided by the Court of Appeals for the Federal Circuit (CAFC) Judge Rader (now Chief Judge).  Sidley Austin Brown & Wood represented Cornell, while HP was represented by the firms DLA Piper and Harter, Secrest & Emery.

The patent at issue covered an “instruction issuing mechanism” that enabled computer microprocessors to work faster by executing multiple instructions simultaneously rather than one at a time - obviously a highly sought after marketing advantage in a market where it seems every new computer has to be faster than the one before it.

The jury found Cornell’s patent valid and infringed and awarded $184 million dollars based on a reasonable royalty determination.  These damages were calculated based on a 0.8% percent royalty applied to a $23 billion dollar royalty base. In a Judgment as a Matter of Law, Judge Rader reduced the royalty base and, using the same royalty rate of 0.8%, he reduced the award to $53 million, giving HP the option of the judgment or a new trial on damages.

Cornell’s expert, Dr. Marion Stewart, had originally attempted to argue that the appropriate royalty base consisted of servers and workstations.  Judge Rader conducted a Daubert hearing during the trial and determined that the patent couldn’t be shown to be the basis for demand of servers and workstations.  Rader ruled that “Dr. Stewart did not supply credible and sufficient economic proof to support application of the entire market value rule. Rather, Dr. Stewart tried to present evidence that would mislead the jury to award damages far in excess of their compensatory purpose.” He allowed Stewart to change his testimony for trial, and Stewart came back with a royalty base consisting of a CPU brick – a combination of a processor, a CPU module, a temperature controlling solution, external cache memory, and a power converter. Dr. Stewart’s trial testimony was ultimately excluded as well, as Rader noted, “on more than one occasion and in contravention of this court's order, Dr. Stewart continued to advise the jury that, in his opinion, server and workstation revenues were the appropriate royalty base.”

Rader concluded that the CPU brick as a royalty base was also overreaching. Rader ultimately ruled that Cornell’s damages evidence failed to meet the requirements of the EMVR. Specifically, that Cornell failed to offer “credible and sufficient economic proof that the patented invention drove demand.” In reducing the royalty base, Rader noted that the CPU processor was the smallest salable unit containing the patented invention and the CPU processor was therefore the appropriate royalty base.

Looking more closely at Rader’s opinion, Rader brings us back to the basics by stating that in order to invoke the EMVR:

  • The patented components must be the basis for customer demand for the entire machine including the parts beyond the claimed invention, Fonar Corp. v. General Electric Co., 107 F.3d 1543, 1552 (Fed. Cir. 1997); AND
  • The individual infringing and non-infringing components must be sold together so that they constitute a functional unit or are parts of a complete machine or single assembly of parts, Paper Converting Machine Co. v. Magna-Graphics Corp., 745 F.2d 11, 23 (Fed. Cir. 1984); AND
  • It is not enough that the infringing and non-infringing parts are sold together for mere business advantage, Rite-Hite v. Kelley, 56 F.3d 1538, 1550 (Fed. Cir. 1995).

Judge Rader acknowledged that Cornell had offered the following arguments in pursuit of proving these points:

  • Cornell relied on internal HP documents predicting that the type of functionality achieved in the patent “would be a competitive requirement;”
  • Cornell attempted to show that customers “chose” HP products incorporating the claimed invention by showing an increase in sales of certain systems incorporating the patent; and
  • Cornell argued that CPU bricks were the more appropriate royalty base rather than processors because the market for processors was small and any extrapolation of processor selling prices would be too great an estimate, rather than a firm calculation which Cornell believed it was able to do with price list data for bricks.

In accordance with his ruling, Rader found fault with each of these arguments.  First, he ruled the internal HP documents weren’t sufficient to prove demand because they were not backed by any sort of customer survey or demand evidence linking customer demand with the claimed invention. Second, where Cornell attempted to show increasing sales due to the patent, the Rader noted that there were other factors driving those sales increases and that the comparisons related to processors and not bricks. Finally, on the third point Rader noted that while bricks were listed in pricing data, the parties had acknowledged that there was no market for individual brick sales, unlike the processors for which there was at least a minimal market.

There is no discussion in the opinion of the reasonableness of the 0.8% royalty rate that was left unchanged by Judge Rader, so it is unclear why the same rate would be applicable to both royalty bases.  We are left only with Rader’s statement, “the jury has articulated what it identified as the appropriate royalty rate, see Verdict Form, D.I. 1029, May 30, 2008 at 6, this court has no reason to disturb that rate simply because it found error in the separately articulated royalty base determination. Because the jury unequivocally communicated its royalty rate decision— an unchallenged decision supported by substantial evidence—upholding the jury's royalty rate determination is the option ‘most faithful to the jury's verdict.’"

Lessons Learned

We see from Rader’s opinion that the EMVR is alive and well because he does apply the rule in his selection of the processor as the royalty base, since the processor included unpatented elements in addition to the patented element. Again Rader’s justification for doing so appears to be simply that the processor was the smallest salable unit containing the patented invention. Generalizing this lesson, we are advised to look at the most concise commercial embodiment of the patented invention when seeking to apply the EMVR.

Second, look for external market evidence linking customer demand to the patented feature. In Cornell, internal infringer correspondence discussing the importance of the patented utility was not enough. In many cases, external market evidence such as customer surveys or correspondence has often been ruled adequate – and this can be costly evidence to gather.

Third, evidence of the incremental value of the patented feature should isolate the patented feature as much as possible. In Cornell, Rader determined too many other factors contributed to HP’s sales to warrant a conclusion on the incremental value of the patented feature. Best case scenario: the patentee or the infringer sells two products where the only difference is the patented feature and the incremental value of the patent can be readily ascertained. As this scenario is difficult to find in practice, sophisticated and more costly statistical analysis may yield better answers.

Finally, ignore the judge’s orders at your peril.  Few orders I’ve read discuss an expert witness by name in an uncomplimentary fashion as much as in Cornell, where Rader mentioned Dr. Stewart no less than 20 times and ultimately excluded his testimony.

These lessons are good ones and I discuss more lessons from other cases in my recent co-authored Business Value Monitor article: Recent Court Opinions Impact Patent Valuation.

I understand HP is appealing this verdict, so the final chapter hasn’t been written. Regardless, as we’ve seen in other CAFC rulings, we can expect more rulings to follow Cornell in their demand for greater evidentiary support to justify the application of the EMVR.

Drew Voth, Director at Grant Thornton LLP, is a Certified Public Accountant, Certified Valuation Analyst, Certified Fraud Examiner and a Certified Insolvency and Restructuring Advisor. Over the past 18 years, he has assisted clients resolve a wide range of intellectual property and other disputes, and has testified in arbitration and at trial on a variety of matters. He has authored numerous continuing education courses and spoken frequently on damages topics. Contact Drew at 206.398.2488 or

blog comments powered by Disqus