Today, a three-judge Federal Circuit panel (Prost (author),  Dyk and Hughes) issued its awaited decision in CSIRO v. Cisco that agreed-in-part and disagreed-in-part with Judge Davis’ damages award based on patents alleged to be essential to the IEEE 802.11 WiFi standard, but which patents did not have any FRAND or other standard-setting obligation (see our July 28, 2014 post on Judge Davis’ decision).  This is an important decision that provides incremental insight into proving and determining a reasonable royalty for a standard essential patent, which includes further insight into the Federal Circuit’s first decision on this issue a year ago in Ericsson v. D-Link that involved a standard essential patent that did have a FRAND obligation under the IEEE 802.11 WiFi standard (see our Dec. 5, 2014 post on the Ericsson v. D-Link decision).

This is an important decision to read directly to catch all the nuances and import of the decision, and the incremental guidance it provides in determining a royalty rate as a matter of patent damages law for past infringement of a patent that is essential to a standard.  A few particularly important points come from the decision.

First, the Federal Circuit soundly rejected as “untenable” the accused infringer’s argument that there is a “rule” that all patent damages methodologies always must start out using the smallest salable patent-practicing unit.  The smallest salable patent practicing unit is a principle that can aid courts to determine if a damages expert’s methodology reliably apportions to the patent only the value that the patented technology provides to the infringing product and not other unpatented features.  But it is not the only approach that may be considered, and different cases present different factual circumstances that could lend themselves to different reliable methodologies.  For example, damages methodologies properly may rely on real-world comparable licenses to reliably apportion value to the patented technology, whether the royalties are based on end products or components thereof.  This decision may very well put to rest arguments that there is some “rule” requiring use of the smallest salable patent-practicing unit or that there is any problem per se in royalties being based on the end product rather than its components.

Second, the Federal Circuit clarified that the need to apportion the value of the patented technology from the value of standardization applies whether or not a standard essential patent is subject to a FRAND or other standard setting obligation.  This is based on the long-standing, fundamental principal that statutory damages for infringement under 35 U.S.C. § 284 must be based on the value of the patented invention and not other unpatented features, whether that’s other unpatented technology in an infringing  product or the value of the patent being essential to a standard.


Patent owner Commonwealth Scientific and Industrial Research Organization (“CSIRO”) is the principal scientific research organization for the Austrialian Federal Government.  The patent-in-suit (U.S. Patent No. 5,487,069 or “the ‘069 Patent”) addresses multipath problems in a wireless local area network.  That technology was incorporated into certain versions of the IEEE 802.11 WiFi standard, including revision “a” adopted in 1999 and revision “g” adopted in 2003.  In December 1998, before IEEE adopted revision “a”, CSIRO provided the IEEE with a letter of assurance that it would license the specific patent-in-suit on RAND-terms if the patent were essential to the 802.11a standard.  IEEE sought additional letters of assurance from CSIRO for later revisions of the standard, but CSIRO declined to provide them.

Radiata Technology License Agreement (TLA).  Shortly after the patent issued, a company called Radiata Communications (“Radiata”)  was formed by the named inventor, CSIRO and others to commercialize the patented technology.  Radiata employed various CSIRO employees as well as another named inventor.  CSIRO entered a Technology License Agreement (TLA) with Radiata in February 1998 that, among other things, had a per-WiFi chip royalty payment, decreasing from 5% royalty per chip to 1% as the volume of licensed chips increased.  In 2001, Cisco acquired Radiata and started paying Radiata’s license fees under the TLA license agreement for Radiata products.  This agreement was renegotiated several times, always keeping the general concept of a per-chip royalty base.

CSIRO’s License Rate Card.  In 2003, CSIRO offered industry participants a license on RAND terms on all versions of the standard (at first indicating that it had agreed with IEEE to do so, but later clarifying there was no RAND obligation).  By June 2004, CSIRO developed a Voluntary Licensing Program offering licenses to the ‘069 Patent under “a flat-fee royalty, charged per end product unit sold” under what it called a “Rate Card” structure.  The lowest royalty rate under this structure was $1.40 to $1.90 per unit.  But CSIRO did not have anyone take a license under this Rate Card.

Cisco’s Offer In Negotiations.  In 2004, CSIRO approached Cisco about licensing the patent under the Rate Card schedule, but Cisco did not accept that offer.  During discussions in 2005, Cisco’s Vice-President of Intellectual Property informally suggested that $0.90 per unite might be an appropriate royalty rate.  That rate was about what Cisco had been paying under the initial TLA agreement after Cisco acquired Radiata.

District Court Litigation.  In July 2011, CSIRO sued Cisco for infringing the patent-in-suit.  Both parties stipulated to a bench trial solely on damages and that Cisco would not challenge the patent’s infringement or validity.  In February 2014, Judge Davis held a four-day bench trial on damages.

CSIRO’s damages model was premised on the profit difference between (1) Cisco products using versions of the 802.11 WiFi standard that incorporated the patented technology (IEEE 802.11 versions a and g) and (2) Cisco products using versions of the WiFi standard that did not use the patented technology (IEEE 802.11 version b).  CSIRO argued that the difference between the two versions primarily was attributable to the patented technology.  This led to CSIRO proposing a volume-tiered royalty ranging from $1.35 to $2.25 per end unit (totaling about $30 million for past infringement).

Cisco’s damages model was premised on the original Radiata TLA, leading to a volume-tiered royalty ranging from $0.03 to $0.37 per WiFi chip (totaling about $1 million for past infringement).

Judge Davis rejected both proposed damages models.  Rather, he created a model premised on CSIRO’s 2004 Rate Card and the rate informally suggested by Cisco’s Vice-President during negotiations with CSIRO.  This led to a “reasonable starting point” for negotiations of a royalty in the range from $0.90 to $1.90 per unit.  Judge Davis then considered various Georgia-Pacific factors to the extent he deemed them appropriate and ultimately concluded they were neutral and, thus, did not require adjusting the royalty rate further.  While he used that royalty rate for the Cisco-branded products, he did downwardly adjust that rate to Cisco’s Linksys-branded products because they had a lower profit margin.

So Judge Davis ultimately awarded a volume-tiered royalty structure that ranged from $0.90 to $1.90 for Cisco-branded products and a slightly lower structure ranging from $0.65 to $1.38 for the Linksys-branded products.  This led to a total damages amount of $16,243,067 for past infringement.

Below  is a table comparing the royalty argued by the parties and that awarded by Judge Davis:

  • CSIRO                 $1.35 to $2.25 per unit ($30 million total)
  • Judge Davis        $0.65 to $1.90 per unit ($16 million total)
  • Cisco                   $0.03 to $0.37 per unit ($1 million total)

Federal Circuit Decision

On appeal to the Federal Circuit, Cisco raised three arguments seeking reversal, alleging that Judge Davis erred by:

  • Not starting with the wireless chip as the royalty base, which was the smallest salable patent practicing unit
  • Not adjusting Georgia-Pacific factors to account for the patent being essential to a standard
  • Not giving credit to Cisco’s TLA evidence.

Smallest Salable Patent-Practicing Unit.  The Federal Circuit started with what it deemed a long-standing rule of apportionment when awarding damages for patent infringement under 35 U.S.C. § 284, stating:

Under § 284, damages awarded for patent infringement must reflect the value attributable to the infringing features of the product, and no more.  This principle—apportionment—is the governing rule where multi-component products are involved.  Consequently, to be admissible, all expert damages opinions must separate the value of the allegedly infringing features from the value of all other features. [internal quotations and citations omitted].

Courts must use their gate-keeping authority to ensure expert testimony “using whatever methodology” is “sufficiently reliable to support a damages award,” because parties have “great financial incentive … to exploit the inherent imprecision in patent valuation.”  The Federal Circuit said that the “essential requirement” for such reliability is apportionment:

And as we have repeatedly held, the essential requirement for reliability under Daubert is that the ultimate reasonable royalty award must be based on the incremental value that the patented invention adds to the end product.  In short, apportionment. [internal quotations and citations omitted]

There may be more than one reliable method to estimate the royalty, since different cases present different facts, but its important that whatever methodology is used is “sufficiently tied to the facts of the case”:

In practice, this means that abstract recitations of royalty stacking theory, and qualitative testimony that an invention is valuable—without being anchored to a quantitative market valuation—are insufficiently reliable.  Where the data used is not sufficiently tied to the facts of the case, a damages model cannot meet the substantive statutory requirement of apportionment of royalty damages to the inventions value. [internal quotations and citations omitted]

The Federal Circuit then discussed the smallest salable patent-practicing unit being one principle that can “aid courts in determining when an expert’s apportionment is reliable.”  There are two justifications for it:

First, where small elements of multi-component products are accused of infringement, calculating a royalty on the entire product carries a considerable risk that the patentee will be improperly compensated for non-infringing components of that product.  Second is the important evidentiary principle that care must be taken to avoid misleading the jury by placing undue emphasis on the value of the entire product.  *** Fundamentally, the smallest salable patent-practicing unit principle states that a damages model cannot reliably apportion from a royally base without that base being the smallest salable patent-practicing unit.  [internal quotations and citations omitted]

But the court found that principle did not apply here, because the district court was not apportioning from a royalty base, but started with real-world negotiations between the parties that “already built in apportionment” based on Cisco’s suggested $0.90 per unit royalty as a lower bound and the $1.90 per unit royalty from CSIRO’s Rate Card as the upper bound:

Because the parties’ discussions centered on a license rate for the ‘069 patent, this starting point for the district court’s analysis already built in apportionment.  Put differently, the parties negotiated over the value of the asserted patent, and no more.  The district court still may need to adjust the negotiated royalty rates to account for other factors, but the district court did not err in valuing the asserted patent with reference to end product licensing negotiations. [internal quotations and citations omitted]

The Federal Circuit soundly rejected as “untenable” Cisco’s suggested “rule” that “would require all damages models to begin with the smallest salable patent practicing unit,” which would conflict with other accepted methodologies such as using comparable licenses:

The rule Cisco advances—which would require all damages models to begin with the smallest salable patent-practicing unit—is untenable.  It conflicts with our prior approvals of a methodology that values the asserted patent based on comparable licenses.  Such a model begins with rates from comparable licenses and then accounts for differences in the technologies and economic circumstances of the contracting parties.  Where the licenses employed are sufficiently comparable, this method is typically reliable because the parties are constrained by the market’s actual valuation of the patent.  Moreover, we held in Ericsson that otherwise comparable licenses are not inadmissible solely because they express the royalty rate as a percentage of total revenues, rather than in terms of the smallest salable unit.  Therefore, adopting Cisco’s position would necessitate exclusion of comparable license valuations that—at least in some cases—may be the most effective method of estimating the asserted patent’s value.  Such a holding would often make it impossible for a patentee to resort to license-based evidence.

Accordingly, we conclude that the district court did not violate apportionment principles in employing a damages model that took account of the parties’ informal negotiations with respect to the end product. [internal quotations and citations omitted]

The Federal Circuit stated in a footnote that the choice of royalty base did not matter in the district court’s analysis in this particular case because it was per unit of end product, which in this case “could equally have represented” a per unit royalty per WiFi chip “without affecting the damages calculation.”

Standardization.  The Federal Circuit agreed with Cisco that the district court erred by not “account[ing] for any extra value accruing to the ‘069 patent from the fact that it is essential to the 802.11 standard.”  Relying on the Ericsson v. D-Link decision, the Federal Circuit stated that there are “unique considerations that apply to apportionment in the context of a standard-essential patent”, and these considerations applied even for standard essential patents that did not have a RAND or other standard-setting obligation (as is the case for the CSIRO patents at issue here).  Damages methodologies applied to standard essential patents (“SEPs”) must “capture the asserted patent’s value resulting … only from the technology’s superiority” and “not from the value added by the standard’s widespread adoption.”

[A] reasonable royalty calculation under § 284 attempts to measure the value of the patented invention.  This value—the value of the technology—is distinct from any value that artificially accrues to the patent due to the standard’s adoption.  Without this rule, patentees would receive all of the benefit created by standardization—benefit that would otherwise flow to consumers and businesses practicing the standard.  We therefore reaffirm that reasonable royalties for SEPs generally—and not only those subject to a RAND commitment—must not include any value flowing to the patent from the standard’s adoption.

The Federal Circuit found that Judge Davis erred by not taking account for standardization but, instead, he “increased the royalty award because the ‘069 patent is essential to the 802.11 standard.”  For example, Judge Davis found that Georgia-Pacific Factors 8-10 favored increasing the royalty given commercial success and widespread adoption of the technology, but “the district court never considered the standard’s role in causing commercial success.”  Further, Judge Davis relied on the rates as a starting point without considering whether such rates “themselves may be impacted by standardization”:

The parties do not dispute that CSIRO actively refused to submit a letter of assurance to the standard-setting body, for later iterations of the 802.11 standard, after the ‘069 patent was locked into the standard.  It seems quite possible, then, that CSIRO’s Rate Card rates attempt to capture at least some value resulting from the standard’s adoption.  CSIRO’s offer was not accepted by a single entity.  On remand, the district court should consider whether the initial rates taken from the parties’ discussions should be adjusted for standardization.

TLA Agreement Between CSIRO and Radiata.  The Federal Circuit found this to be a fact sensitive issue whether Judge Davis erred in not giving more weight to the TLA, and ultimately found that he had erred and should reconsider the issue on remand.  Among other things, the agreement “is the only actual royalty agreement between Cisco and [CSIRO]” and “it is contemporaneous with the hypothetical negotiation.”