Broadcom Corp. v. Qualcomm

United States Court of Appeals for the Federal Circuit
2008-1199, -1271, -1272

BROADCOM CORPORATION,

Plaintiff-Appellee,

v.

QUALCOMM INCORPORATED,

Defendant-Appellant.

William F. Lee, Wilmer Cutler Pickering Hale and Dorr LLP, of Boston,
Massachusetts, argued for plaintiff-appellee. With him on the brief were Richard W.
O’Neill, Joseph J. Mueller, and Lauren B. Fletcher; and James L. Quarles, III, of
Washington, DC. Of counsel was Heath A. Brooks, of Washington, DC.

Evan R. Chesler, Cravath, Swaine & Moore LLP, of New York, New York, argued
for defendant-appellant. With him on the brief were Richard J. Stark and Andrei
Harasymiak. Of counsel on the brief were Carter G. Phillips and Stephen B. Kinnaird,
Sidley Austin LLP, of Washington, DC, and Richard T. Mulloy and Stanley J.
Panikowski, DLA Piper US LLP, of San Diego, California.

Richard McMillan, Jr., Crowell & Moring, LLP, of Washington, DC, for amicus
curiae, Sprint Nextel Corporation. With him on the brief were Kathryn L. Clune, Brian M.
Koide, and Nathaniel Grow.

Appealed from: United States District Court for the Central District of California

Judge James V. Selna
 United States Court of Appeals for the Federal Circuit

2008-1199, -1271, -1272

BROADCOM CORPORATION,

Plaintiff-Appellee,
v.

QUALCOMM INCORPORATED,

Defendant-Appellant.

Appeal from the United States District Court for the Central District of California in case
no. 05-CV-467, Judge James V. Selna.

__________________________

DECIDED: September 24, 2008
__________________________

Before LINN, FRIEDMAN, and PROST, Circuit Judges.

LINN, Circuit Judge.

Qualcomm Incorporated (“Qualcomm”) appeals from a jury’s determination that

Qualcomm infringed U.S. Patents No. 6,847,686 (“the ’686 patent”), No. 5,657,317 (“the

’317 patent”), and No. 6,389,010 (“the ’010 patent”), owned by Broadcom Corporation

(“Broadcom”). Qualcomm also appeals from the district court’s issuance of a

permanent injunction against Qualcomm. Because the district court erred in its

construction of claim 3 of the ’686 patent, we reverse the jury’s determination of

infringement of that patent and conclude that claim 3, as properly construed, is invalid.

Because the district court did not err in construing the claims of the ’317 patent, and

because substantial evidence supports the jury’s determinations of infringement and
validity of the ’317 and ’010 patents, we affirm the judgment of infringement of the ’317

and ’010 patents, and the injunction as it pertains to those patents.

I. BACKGROUND

Broadcom and Qualcomm compete in the market for chipsets used in mobile

radio devices such as cell phone handsets. The technology at issue in this appeal

relates to wireless voice and data communications on cellular telephone networks. The

relevant technology is currently found in so-called third-generation (“3G”) baseband

processor chips. Baseband processor chips enable a cell phone’s basic communication

functions, along with other features such as graphics, multimedia, data transfer, and

custom applications. The 3G chips sold by Broadcom, Qualcomm, and others replace

older and less capable second-generation (“2G”) chips, which include code division

multiple access (“CDMA”) chips and global system for mobile communications (“GSM”)

chips. The 3G CDMA replacement is known as CDMA2000, while the 3G GSM

replacement is known as wideband CDMA (“WCDMA”). These 3G technologies are

generally incompatible with each other, and thus both cell phone handsets and cell

phone service provider networks are designed to work with only one of these two

competing standards. Both standards, however, provide enhanced functionality over

their 2G predecessors, particularly in the area of multimedia and multiple network

products.

The products accused of infringing Broadcom’s patents include baseband chips

designed to work in cell phones in conjunction with 3G networks, including both

CDMA2000 and WCDMA, although this appeal primarily relates to CDMA2000 chips.

Also at issue is Qualcomm’s software, which Qualcomm has licensed exclusively to

2008-1199, -1271, -1272 2
Sprint. QChat allows the use of push-to-talk (“PTT”) technology much like a walkie-

talkie on CDMA2000 networks. Although Broadcom markets 3G chips, it does not

currently sell any CDMA2000 chips, it has not sold any WCDMA chips for use in United

States cell phones, and it does not offer a chip implementing a PTT feature.

Broadcom’s ’686 patent relates to video compression technology on cell phone

devices. Claim 3, the only claim at issue, depends from claim 1, and is directed to a

“digital signal processor” (“DSP”). Broadcom accused Qualcomm’s WCDMA and

CDMA2000 baseband processor chips of infringing this claim. The ’317 patent relates

to technology allowing cell phones to simultaneously participate on multiple wireless

networks using a single transceiver. The relevant claims on appeal are directed to

“radio units” having “transceivers” capable of “simultaneous” participation on multiple RF

networks. Broadcom asserted that Qualcomm’s CDMA2000 chips infringed these

claims by interfacing with both the 1x network (for traditional voice communications) and

the EV-DO network (for data and related applications including PTT functionality in

QChat). Finally, the ’010 patent claims a “telephone” having circuitry allowing it to

“selectively couple” to two networks having different bandwidth characteristics.

Broadcom asserted the several claims of the ’010 patent against Qualcomm’s

CDMA2000 chips implementing the QChat PTT feature. Broadcom argued that that

traditional voice calls couple to the telephone network, while calls initiated with the

QChat feature are routed through the Internet, a distinct network.

The district court issued a claim construction order prior to trial in which it

construed the contested terms of Broadcom’s patents pursuant to Markman v.

Westview Instruments, Inc., 517 U.S. 370, 372 (1996). Broadcom Corp. v. Qualcomm

2008-1199, -1271, -1272 3
Inc., No. 05-CV-467 (C.D. Cal. Sept. 11, 2006) (“Claim Construction Order”). Following

a trial, the jury found that Qualcomm directly infringed and induced infringement of claim

3 of the ’686 patent and claims 1, 6, 9, and 12 of the ’317 patent, either literally or under

the doctrine of equivalents; that Qualcomm directly infringed, induced infringement of,

and contributed to the infringement of claims 1, 2, 3, and 7 of the ’010 patent, either

literally or under the doctrine of equivalents; and that all of these claims were not invalid.

The jury further found that Qualcomm had willfully infringed all three patents in suit. It

awarded damages of approximately $20 million. Qualcomm filed post-trial motions for

judgment as a matter of law under Rule 50(b) of the Federal Rules of Civil Procedure or

a new trial under Rule 59(a). The district court denied relief, denying Qualcomm’s

motions in their entirety. Broadcom Corp. v. Qualcomm Inc., No. 05-CV-467 (C.D. Cal.

Aug. 10, 2007) (“JMOL Opinion”).

Ten days after the district court’s denial of Qualcomm’s post-trial motions, this

Court released its decision in In re Seagate Technology, LLC, 497 F.3d 1360 (Fed. Cir.

2007) (en banc). Given the potential impact of Seagate on the willfulness determination

previously made by the jury, the district court sua sponte invited a motion for

reconsideration of its denial of Qualcomm’s request for a new trial on willfulness and of

its award of enhanced damages. Upon considering the arguments presented, the

district court vacated the willfulness verdict and directed Broadcom to file an election

“indicating whether it wishe[d] to accept the liability and damages verdicts or proceed to

a new trial on all issues, including willfulness.” Broadcom Corp. v. Qualcomm Inc., No.

05-CV-467, slip op. at 9 (C.D. Cal. Nov. 21, 2007). Broadcom elected to accept the

liability and damage verdicts rather than proceed to a new trial on all issues.

2008-1199, -1271, -1272 4
 The district court then held a bench trial on the topic of injunctive relief and

subsequently entered a permanent injunction against Qualcomm on all three patents,

although it provided “sunset” provisions allowing continued sales pursuant to a

mandatory royalty through January 31, 2009. Broadcom Corp. v. Qualcomm Inc., No.

05-CV-467 (C.D. Cal. Dec. 31, 2007) (“Injunction Opinion”). On February 5, 2008, the

district court issued an amended injunction, Broadcom Corp. v. Qualcomm Inc., No. 05-

CV-467 (C.D. Cal. Feb. 5, 2008) (“First Amended Injunction”), and on March 13, 2008, it

again amended and re-issued the injunction, Broadcom Corp. v. Qualcomm Inc., No.

05-CV-467 (C.D. Cal. Mar. 13, 2008) (“Second Amended Injunction”). Finally, on March

24, 2008, the district court entered judgment pursuant to Federal Rule of Civil

Procedure 54(b). Broadcom Corp. v. Qualcomm Inc., No. 05-CV-467 (C.D. Cal. Mar.

24, 2008).

Qualcomm appealed from the First Amended Injunction, the Second Amended

Injunction, and the Rule 54(b) judgment entering the jury’s infringement verdicts. We

granted unopposed motions to consolidate these three appeals on April 1, 2008, and

April 11, 2008. We have jurisdiction over the appeal from the Rule 54(b) judgment

pursuant to 28 U.S.C. § 1295(a)(1), and we have jurisdiction over the appeals from the

interlocutory injunctions pursuant to 28 U.S.C. § 1292(c)(1).

II. DISCUSSION

2008-1199, -1271, -1272 5
 Qualcomm presents numerous arguments regarding claim construction,

infringement, and validity, as well as various contentions regarding the necessity of a

new trial and the propriety of the district court’s permanent injunction. We address each

in turn.

A. The ’686 Patent

Claim 3, the only asserted claim of the ’686 patent, depends from claim 1, the

patent’s only independent claim, which recites:

A digital signal processor for processing a multiple frame video digital
signal, comprising:
a DSP controller,
a plurality of processing units connected to said DSP controller for
processing said multiple frame video digital signal; and
at least one storage unit, wherein each of said processing units is
connected to at least one of said at least one storage units,
said DSP controller controlling said plurality of processing units,
wherein said DSP controller, said plurality of processing units, and
said at least one storage unit are on a single chip.

Dependent claim 3 additionally recites “[t]he digital signal processor according to claim

1 wherein each of said processing units operates according to a different program

command.” Although the claims make no reference to a “global controller,” the district

court construed the claim term “DSP controller controlling said plurality of processing

units” as follows: “The DSP controller, either independently or under the direction of a

global controller, distributes control instructions to be executed by the plurality of

processing units. A global controller is required.” J.A. at 2251 (emphasis added). It is

undisputed that the Texas Instruments TMS320C80 chip and its documentation (“C80

chip”) would invalidate claim 3 as an anticipatory reference if the claim is construed not

to include a “global controller” limitation.

2008-1199, -1271, -1272 6
 Qualcomm contends that the district court improperly imported the “global

controller” limitation into claim 1 (and, thus, claim 3) of the ’686 patent and that the C80

chip anticipates this claim. Qualcomm argues that the claim language is directed solely

to DSPs, and that the specification and figures of the patent clearly distinguish between

DSPs and global controllers. Qualcomm had argued that the DSP controller term

should be construed as “[u]nder the direction of a global controller, the DSP controller

controls units that process data.” J.A. at 550. Broadcom responds that the district court

merely construed the claim terms in light of the specification, which discusses how the

“[g]lobal controller . . . controls and schedules . . . the digital signal processor,” ’686

patent col.7 ll.9-11, and that to ignore the “global controller” would render claim 1 invalid

even with respect to generic prior art discussed in the specification. Claim construction

is a question of law that we review de novo. Cybor Corp. v. FAS Techs., Inc., 138 F.3d

1448, 1454 (Fed. Cir. 1998) (en banc).

We agree with Qualcomm. There is no basis for importing the “global controller”

limitation into claim 1. The claims make no reference to an external global controller,

and we find no justification for adding limitations related to the more general video

encoding device discussed in the specification when the claims are written to

encompass only a specific element of that device. The claims recite only three DSP

elements, including a DSP controller, a plurality of processing units, and at least one

storage unit. ’686 patent col.16 ll.6-21. The specification similarly provides that the

DSP includes only these three components. E.g., id. col.4 ll.44-49 (“The digital signal

processor includes a DSP controller, a plurality of processing units, . . . and at least one

storage unit. . . . The DSP controller controls the plurality of processing units.”). The

illustrations provided in the specification further bolster our conclusion that the global

2008-1199, -1271, -1272 7
controller is separate and distinct from the DSP. For example, Figure 7, reproduced

below, places the global controller outside of the lines defining the DSP.

Figure 8 similarly illustrates the global controller as lying outside the bounds of the DSP.

Figures 3 and 4 also identify the global controller and the digital signal processor as

discrete and separate components.

While the specification does contemplate the possibility of a global controller, the

claims of the ’686 patent are directed solely to the DSP. As this court has noted, “each

claim does not necessarily cover every feature disclosed in the specification. When the

claim addresses only some of the features disclosed in the specification, it is improper

to limit the claim to other, unclaimed features.” Ventana Med. Sys., Inc. v. Biogenex

Labs., Inc., 473 F.3d 1173, 1181 (Fed. Cir. 2006). While the ’686 patent claims only the

DSP, it is a continuation of another application—sharing the same specification—now

issued as U.S. Patent No. 6,385,244 (“the ’244 patent”). The ’244 patent has two

independent claims. One claims “a video encoding system including . . . a video

encoding device . . . the video encoding device comprising: . . . a global controller . . .

2008-1199, -1271, -1272 8
[and] a digital signal processor, connected to said global controller.” ’244 patent

claim 1. The other claims “[a] video camera comprising: . . . a video encoding device . .

. comprising . . . a global controller . . . [and] a digital signal processor, connected to

said global controller.” Id. claim 7. While the global controller limitation influences the

scope of the ’244 patent claims, we find no basis for importing it to the DSP claims of

the ’686 patent.

Broadcom also contends that the claim cannot be construed without the global

controller limitation because claim 1 otherwise would be “self-invalidating” based on

generic prior art disclosed in the specification. Broadcom points to Figure 2 of the ’686

patent—labeled “PRIOR ART”—to support this argument.

“A patent is invalid as anticipated if every limitation in a claim is found in a single

prior art reference.” Nystrom v. TREX Co., 424 F.3d 1136, 1149 (Fed. Cir. 2005).

Because Broadcom has not demonstrated that “every limitation” of claim 1 is found in

this reference, its self-invalidating argument must fail. For example, Broadcom does not

even allege that the generic prior art DSP referenced in the patent is implemented “on a

single chip” as required by the claims. Moreover, “[w]hile we have acknowledged the

2008-1199, -1271, -1272 9
maxim that claims should be construed to preserve their validity, we have not applied

that principle broadly . . . . Instead, we have limited the maxim to cases in which the

court concludes, after applying all the available tools of claim construction, that the

claim is still ambiguous.” Phillips v. AWH Corp., 415 F.3d 1303, 1327 (Fed. Cir. 2005)

(en banc) (internal citation and quotation marks omitted); see also Tate Access Floors,

Inc. v. Interface Architectural Res., Inc., 279 F.3d 1357, 1372 (Fed. Cir. 2002) (noting

that “where claim language is clear we must accord it full breadth even if the result is a

claim that is clearly invalid”).

Accordingly, we modify the district court’s claim construction to remove the

requirement of a “global controller,” reverse the jury verdict of infringement as to the

’686 patent, and hold claim 3 invalid as anticipated by the C80 chip. See Akamai

Techs., Inc. v. Cable & Wireless Internet Servs., Inc., 344 F.3d 1186, 1194-95 (Fed. Cir.

2003).

B. The ’317 Patent

The jury found that Qualcomm infringed claims 1, 6, 9, and 12 of the ’317 patent.

These claims do not meaningfully differ for purposes of this appeal. Claim 1 recites,

with disputed terms emphasized:

A radio unit for operation in a communication system having a plurality of
RF communication networks comprising:
a transceiver capable of participating on the plurality of RF
communication networks;
a memory device which stores a plurality of communication
protocols, each communication protocol governing radio
operation on one of the plurality of RF communication
networks;
a control processor coupled to the transceiver and the memory
device, the control processor selecting from the memory
device ones of the plurality of communication protocols to
enable the transceiver to simultaneously participate on

2008-1199, -1271, -1272 10
 corresponding ones of the plurality of RF communication
networks; and
the control processor managing the simultaneous use by the
transceiver of the selected ones of the plurality of
communication protocols.

Qualcomm argues that the district court improperly construed the “simultaneously

participate” limitation in all of the relevant claims of the ’317 patent, that there is no

evidence of infringement under a correct construction, and that the ’317 patent is invalid

under the district court’s claim construction. It also argues that the district court erred in

not construing the “networks” limitation, and that substantial evidence does not support

the jury’s infringement verdict under a correct construction.

1. “Simultaneously Participate”

Claims 1, 6, 9, and 12 of the ’317 patent all require a radio unit with a transceiver

that can “simultaneously participate” on two or more wireless networks. The district

court construed this claim term as “[t]aking part in communications with two or more

networks either actively or in sleep-mode during the same period of time.” Claim

Construction Order at 12 (emphasis added). Qualcomm argues, as it did to the district

court, that “simultaneously participate” requires the claimed transceiver to be capable of

communicating on multiple networks at the same instant in time. Broadcom defends the

district court’s claim construction as supported by the specification, which it contends

distinguishes between “simultaneous” and “fully simultaneous.”

The district court reasoned that “because the ’317 patent claims a radio unit with

a single transceiver, and because a single transceiver cannot achieve full

communication with two networks at the same time, Qualcomm’s construction could

mean that the patent would not perform its stated function. Accordingly, the Court

2008-1199, -1271, -1272 11
agree[d] with Broadcom that the construction ‘during the same period of time’ most

accurately construes the meaning of the term simultaneous.” Claim Construction Order

at 12; see Cordis Corp. v. Medtronic Ave, Inc., 511 F.3d 1157, 1174 (Fed. Cir. 2008)

(noting that “a construction that renders the claimed invention inoperable should be

viewed with extreme skepticism”).

The claim language itself provides little insight as to the meaning of

“simultaneous,” and each party argues that the specification supports its own

interpretation. Qualcomm contends that the specification distinguishes between

“simultaneous participation,” for communications taking place at the same instant in

time, and “multiple participation,” for interleaved communications taking place over the

same period of time. See, e.g., ’317 patent col.12 ll.42-46 (“Although the use of a

plurality of built-in radio transceivers could be used so as to permit simultaneous

participation by a single device, factors of cost, size, power and weight make it desirable

to only build-in a single radio transceiver capable of multiple participation.”). Broadcom

argues that the specification distinguishes between “simultaneous” or “multiple”

participation and “fully simultaneous” participation, and that only the latter denotes non-

interleaved communications taking place at the same instant in time. See, e.g., id.

col.13 ll.1-4 (“In communication environments wherein fully simultaneous participation

does not exist or is not desired, transmitter circuitry might be shared for participation in

both [networks].”). We do not find these arguments to be helpful because the use of

these terms in the specification is inconsistent at best.

We turn then to Figure 34 of the ’317 patent, as reproduced below, which has

been at the center of much debate, both at the district court and on appeal.

2008-1199, -1271, -1272 12
The block labeled 3273 represents an instance of two networks, represented as

timelines 3253 and 3255, seeking to communicate with the same device at the same

time. The specification discussion pertaining to block 3273 provides the following:

“Block 3273 . . . illustrates a situation where the master encounters a communication

conflict . . . . If the master has two radio transceivers, the master can service both

networks. If, however, the master only has one radio transceiver, the master chooses

to service one network based on network priority considerations.” Id. col.53 ll.42-50.

While this language indicates that multiple transceivers may be used to avoid such

conflict (and the attendant interleaving process), the specification as a whole repeatedly

clarifies that the invention is directed to wireless communications solutions utilizing a

single transceiver. For example, in the section labeled “Technical Field,” the

specification states that “[t]he present invention relates generally to local area networks

used for transmitting and receiving information and more particularly to a singular radio

using multiple communication protocols for servicing corresponding multiple radio local

2008-1199, -1271, -1272 13
area networks.” Id. col.1 l.65 – col.2 l.3 (emphasis added). Moreover, as the district

court noted, “both parties concede that the patent discloses a device with one

transceiver.” Claim Construction Order at 12. Based on this context, we agree with the

district court that the claim term “simultaneous participation” refers to interleaved

communications.

Because Qualcomm only argues that there can be no infringement under its

proposed construction, but does not argue non-infringement under the district court’s

construction, we affirm the jury’s infringement verdict.

Qualcomm further argues, however, that even under the district court’s

construction of “simultaneously participate,” it introduced a prior art reference that

anticipates the relevant claims of the ’317 patent. That reference, U.S. Patent No.

5,550,895 (“Burson”), discloses a system allowing a telephone to operate in both

cordless and cellular networks. While the parties dispute whether or not this reference

discloses “simultaneous participation” in these different networks, Broadcom additionally

argued at trial that it could antedate the Burson reference, and thus that Burson is not

prior art capable of anticipating the ’317 patent. Specifically, Broadcom argued that the

September 1989 lab notebook of a deceased inventor, Ronald Mahany, demonstrated

conception of the claimed invention, and that this invention was constructively reduced

to practice in May 1993 and November 1993 patent applications. If true, these facts

permit Broadcom to antedate the Burson reference, which was filed in December

1993—subsequent to both the lab notebook and the 1993 patent applications. See

generally Mahurkar v. C.R. Bard, Inc., 79 F.3d 1572, 1577 (Fed. Cir. 1996). Having

been presented with the deceased inventor’s lab notebook, the patent applications, and

2008-1199, -1271, -1272 14
the testimony of a co-inventor, a coworker, and Broadcom’s expert, the jury rejected

Qualcomm’s invalidity contentions.

Qualcomm argues that neither the lab notebook nor the 1993 patent applications

disclose every limitation of the asserted claims. Qualcomm relies on our recent holding

in PowerOasis, Inc. v. T-Mobile USA, Inc., 522 F.3d 1299, 1305 (Fed. Cir. 2008)

(holding that once a defendant established prior art anticipating the asserted claims,

“the burden was on [the patentee] to come forward with evidence to the contrary”), and

argues that Broadcom failed to carry its burden at trial in demonstrating entitlement to a

filing date prior to that of the Burson reference. We disagree. The record shows that

Broadcom’s expert provided detailed testimony at trial explaining his basis for

concluding that each limitation of the accused claims was present in the 1989 lab

notebook, and that he had gone through the same analysis with respect to the 1993

patent applications. See J.A. at 1507-11. Although Qualcomm focuses on the 1993

patent applications on appeal, it failed to ask Broadcom’s expert any questions about

these applications at trial, and its own expert did not testify on the subject. Indeed, in

rejecting Qualcomm’s post-trial motions, the district court noted that Broadcom’s

witnesses “presented an unrebutted picture of Mahany’s conception at least as early as

September 1989 based on Mahany’s notebooks and constructive reductions to practice

in May 1993.” JMOL Opinion at 9 (emphasis added). Therefore, based on the

unrebutted evidence of record, we affirm the jury’s verdict that the Burson reference

does not invalidate the ’317 patent as an anticipatory reference. See Johnston v. IVAC

Corp., 885 F.2d 1574, 1581 (Fed. Cir. 1989) (noting that “[a]ttorneys' argument is no

substitute for evidence”).

2. “Networks”

2008-1199, -1271, -1272 15
 Qualcomm further argues that the district court erroneously failed to construe the

term “networks” and thus left an issue of claim construction to the jury. See O2 Micro

Int’l Ltd. v. Beyond Innovation Tech. Co., 521 F.3d 1351, 1362 (Fed. Cir. 2008) (“When

the parties present a fundamental dispute regarding the scope of a claim term, it is the

court’s duty to resolve it.”). Qualcomm contends that “network” should be construed as

“a plurality of network devices,” rather than “protocols,” and that no substantial evidence

supports a finding of infringement because 1x and EV-DO are not separate networks,

but merely distinct protocols. Broadcom responds that Qualcomm waived this argument

by not presenting it to the district court until after the jury had returned a verdict.

Qualcomm first raised this argument in its post-trial motions, and the district court held

that “Qualcomm waived any claimed error by virtue of its failure to offer a proposed

construction at or prior to trial.” JMOL Opinion at 2 n.1 (citing Abbott Labs. v. Syntron

Bioresearch, Inc., 334 F.3d 1343, 1352 (Fed. Cir. 2003)).

We agree that Qualcomm cannot be allowed to create a new claim construction

dispute following the close of the jury trial. “[A] party may not introduce new claim

construction arguments on appeal or alter the scope of the claim construction positions

it took below. Moreover, litigants waive their right to present new claim construction

disputes if they are raised for the first time after trial.” Conoco, Inc. v. Energy & Envtl.

Int’l, L.C., 460 F.3d 1349, 1358-59 (Fed. Cir. 2006) (internal citations omitted).

Qualcomm’s eleventh-hour attempt to litigate a newly minted claim construction

controversy falls squarely within our holding in Eli Lilly & Company v. Aradigm

Corporation, where a party “never requested that the district court construe any terms in

[the relevant claim] and never offered a construction of [that claim],” but rather “[o]nly

2008-1199, -1271, -1272 16
after the presentation of all of the evidence to the jury . . . even suggest[ed] that claim

construction might be helpful to determine the proper scope of the claimed invention.”

376 F.3d 1352, 1360 (Fed. Cir. 2004). As in Eli Lilly, we hold that Qualcomm has

waived its right to request a construction of “networks” and that Qualcomm has thereby

implicitly conceded that the meanings of “networks” is clear and not in need of

construction. Id. Qualcomm’s reliance on our recent holding in O2 Micro is misplaced.

Unlike in O2 Micro, where the parties disputed the proper construction of a term at a

pre-trial Markman hearing, 521 F.3d at 1362, Qualcomm here has failed to offer its

proposed construction of “networks” at or prior to trial, and we reject such arguments

raised for the first time after the jury verdict.

C. The ’010 Patent

Qualcomm also argues that substantial evidence does not support the jury’s

verdict of infringement with respect to the ’010 patent. The jury found that Qualcomm

directly infringed, induced infringement of, and contributed to the infringement of claims

1, 2, 3, and 7 of the ’010 patent. Claim 1 is representative and recites, with the relevant

portion emphasized:

A telephone for use with a first network having a predetermined allocated
bandwidth and a second network having a variable bandwidth, the
telephone comprising:
an interface circuit that selectively couples to the first and second
networks, the first and second networks operating
substantially independent of one another;
a user interface;
processing circuitry, coupled to the user interface and the interface
circuit, that directs the interface circuit to select one of the
first and second networks;
a microphone capable of being selectively coupled to the first and
second networks via the interface circuit; and
an analog to digital converter that is coupled by the interface circuit
between the microphone and the second network under the

2008-1199, -1271, -1272 17
 control of the processing circuitry when the second network
is selected;
wherein the processing circuitry responds to the user interface to
determine which of the first and second networks to select.

The accused products include CDMA2000 cell phones with the QChat PTT

feature. Using this technology, cell phone users can choose to either place a regular

voice call, which is routed to the public switched telephone network (“PSTN”) through

the 1x network, or they may use the PTT—i.e., QChat—feature, which routes calls

through the Internet using the EV-DO network when the user pushes a separate QChat

button on the phone. This QChat PTT feature allows the cell phone to be used like a

walkie-talkie. At trial, Broadcom successfully argued that these facts demonstrate that

users of these phones selectively couple to the PSTN and the Internet—the “first and

second networks” for purposes of this patent.

On appeal, Qualcomm argues that allegedly infringing cell phones using the

QChat software do not “selectively couple” to either the PSTN or to the Internet because

these phones do not “couple” to any network through a physical electrical connection,

but rather communicate wirelessly and exclusively with the CDMA2000 infrastructure

through either the 1x or EV-DO protocols. According to Qualcomm, it is then the

CMDA2000 network equipment that “selectively couples” with either the PSTN or the

Internet. It contends that because the claims require “an interface circuit that selectively

couples” to both networks, and because the circuits inside the cell phones only prepare

data and voice packets for transmission, there can be no infringement.

Broadcom responds that the circuitry in the phone is the beginning of the

electrical pathway to either network (the PSTN or the Internet) and that the presence of

intermediaries between the cell phone and the destination networks should not affect

2008-1199, -1271, -1272 18
infringement because the ’010 patent does not require a direct connection. It further

notes that the patent itself explicitly contemplates the use of wireless networks. See,

e.g., ’010 patent col.1 ll.41-44 (noting that “this invention relates to the intelligent routing

of packetized voice communication between telephones and radio terminals through

wireless and hardwired channels in a data processing network” (emphasis added)); id.

col.2 ll.52-56 (“An object of the invention is to provide a method and apparatus wherein

seamless voice and data communication is provided among both roaming devices

within wireless portions of a communication network and stationary devices within

hardwired portions of the network.” (emphasis added)). Even if the claims required a

direct or physical connection, however, Broadcom also argues that it presented

substantial evidence of infringement under the doctrine of equivalents.

The term “selectively couples” was not construed by the district court because

the parties agreed to let the ordinary meaning control. Infringement was presented to a

jury, whose verdict we review for substantial evidence. See Dawn Equip. Co. v. Ky.

Farms Inc., 140 F.3d 1009, 1014 (Fed. Cir. 1998). “In reviewing factual issues for

substantial evidence, the inquiry is whether a reasonable jury, given the record before it

viewed as a whole, could have arrived at the conclusion it did.” Id. Here, the jury was

presented with more than adequate evidence upon which it could have concluded that

phones incorporating Qualcomm’s QChat software included “an interface circuit that

selectively couples” to either the PSTN or the Internet. Undisputed testimony at trial

established that phones using the QChat software distinguish between digitalized voice

data being sent to the PSTN and to the Internet by packaging data into “voice frames”

for traditional phone calls over the PSTN or by packing this data into “voice packets” for

transmission over the Internet. E.g., J.A. at 2106-07. It is further undisputed “that the

2008-1199, -1271, -1272 19
chip differentiates between voice frames and voice packets in determining which

network to communicate with.” JMOL Opinion at 15. Qualcomm’s own expert admitted

that “[t]here is an interface circuit inside the phone, so the interface circuit inside the

phone itself does this framing or packetizing, and sends it out over a single air interface

using two different protocols.” J.A. at 2107. Broadcom’s expert explained to the jury

that “[s]elective coupling . . . doesn’t mean that it has to be directly connected. . . . It

just means that the coupling is electrically prepared within that interface circuit so that

by the time the work of the network begins they can go to a proper destination.” Id. at

1393-94.

Although Broadcom’s expert stated that “selectively couple” means “connected

electrically,” that is, that “[t]here is an electrical pathway to either the first or the second

network,” id. at 1373, he also testified that “the coupling . . . doesn’t mean that it has to

attach the [sic] connect physically right next to each other,” id. at 1397. He further

testified that “the interface circuit is within the phone” and that it “is electrically coupled

to the first and the second network” where these networks are the PSTN and the

Internet. Id. Taken as a whole, this record provides substantial evidence to support a

reasonable jury’s conclusion that Qualcomm infringed the ’010 patent.

In so holding, we reject Qualcomm’s additional contentions that Broadcom has

presented a post-trial claim construction in contravention of our holding in Hewlett-

Packard Company v. Mustek Systems, Incorporated, 340 F.3d 1314, 1320-21 (Fed. Cir.

2003). As we discussed in connection with the “network” limitation of the ’317 patent,

supra, we agree that “where the parties and the district court elect to provide the jury

only with the claim language itself, and do not provide an interpretation of the language

in the light of the specification and the prosecution history, it is too late at the JMOL

2008-1199, -1271, -1272 20
stage to argue for or adopt a new and more detailed interpretation of the claim language

and test the jury verdict by that new and more detailed interpretation.” Hewlett-Packard,

340 F.3d at 1321. Unlike the patentee in Hewlett-Packard, however, Broadcom has not

argued for a new claim interpretation. The evidence shows that users of the QChat

feature press a specific button for the express purpose of establishing a PTT connection

with another user. In doing so, QChat users expressly choose whether to connect

through the Internet using the PTT feature, or to place traditional voice calls through the

PSTN. Qualcomm has not provided convincing evidence that the ordinary meaning of

“selectively couples” requires a direct or physical connection, and the jury was entitled

to credit the substantial evidence discussed above, as general as it may be, in

rendering a verdict in favor of Broadcom.

D. New Trial

Qualcomm argues that the pre-Seagate jury instructions, which prompted the

district court to set aside the willfulness verdict based essentially on Qualcomm’s not

having obtained non-infringement opinion letters, also require us to overturn the indirect

infringement verdicts predicated on specific intent necessary to find inducement under

DSU Medical Corporation v. JMS Company, 471 F.3d 1293 (Fed. Cir. 2006) (en banc in

relevant part). Qualcomm contends that it did not obtain non-infringement opinions

because it had procured invalidity opinions for each relevant patent, although it elected

not to waive attorney-client privilege and rely on those opinions. Qualcomm also argues

that the indirect infringement verdicts are unsupported by substantial evidence. In the

event we accept these contentions, Qualcomm further asserts that we must vacate the

damages awards related to the ’317 and ’010 patents because the damages were

based solely on indirect infringement activities. Finally, it argues that because damages

2008-1199, -1271, -1272 21
and infringement issues have interwoven factual underpinnings, retrial should be

permitted with respect to all issues.

Broadcom responds that the district court emphasized the same DSU specific

intent standard on which Qualcomm relies, and that the disputed instructions regarding

opinion-of-counsel letters properly identified such evidence as one factor to consider

within the totality of the circumstances. Broadcom contends that opinion-of-counsel

letters remain relevant to the intent inquiry of our inducement precedent, and that

Seagate addressed neither the admissibility of evidence concerning an alleged

infringer’s failure to obtain non-infringement opinions, nor the standard for establishing

intent to induce infringement. Broadcom further argues that circumstantial evidence in

the record provides substantial evidence to support the jury’s verdict of induced

infringement.

“Legal standards given to the jury are reviewed without deference to the trial

court, whereas the jury’s resolution of factual disputes are reviewed to determine

whether there is substantial evidence to support them.” Dawn Equip., 140 F.3d at 1014.

“This court reviews the legal sufficiency of jury instructions on an issue of patent law

without deference to the district court. This Court reviews jury instructions in their

entirety and only orders a new trial when errors in the instructions as a whole clearly

mislead the jury.” DSU, 471 F.3d at 1304 (internal quotation marks and citation

omitted). We turn first to the legal standard.

“In order to prevail on an inducement claim, the patentee must establish ‘first that

there has been direct infringement, and second that the alleged infringer knowingly

induced infringement and possessed specific intent to encourage another’s

2008-1199, -1271, -1272 22
infringement.’” ACCO Brands, Inc. v. ABA Locks Mfr. Co., 501 F.3d 1307, 1312 (Fed.

Cir. 2007) (quoting Minn. Mining & Mfg. Co. v. Chemque, Inc., 303 F.3d 1294, 1304-05

(Fed. Cir. 2002)). The relevant standard for establishing the intent element of

inducement was clarified by this Court in DSU, 471 F.3d at 1304-06 (en banc in relevant

part). In DSU, we upheld a jury instruction providing that “[t]he defendant must have

intended to cause the acts that constitute the direct infringement and must have known

or should have known tha[t] its action would cause the direct infringement.” Id. at 1305.

In so doing, we held that “inducement requires evidence of culpable conduct, directed to

encouraging another’s infringement, not merely that the inducer had knowledge of the

direct infringer’s activities.” Id. at 1306 (citing Metro-Goldwyn-Mayer Studios, Inc. v.

Grokster, Ltd., 545 U.S. 913, 936-37 (2005)). We further noted that “[t]he plaintiff has

the burden of showing that the alleged infringer’s actions induced infringing acts and

that he knew or should have known his actions would induce actual infringements.”

DSU, 471 F.3d at 1306 (quoting Manville Sales Corp. v. Paramount Sys., Inc., 917 F.2d

544, 553 (Fed. Cir. 1990)).

In this case, the district court instructed the jury, inter alia, as follows:

In order to establish active inducement of infringement, it is not sufficient
that the company that is allegedly induced to infringe itself directly
infringes the claim. Nor is it sufficient that QUALCOMM was aware of the
act(s) that allegedly constitute the direct infringement. Rather you must
find that QUALCOMM specifically intended to cause direct infringement of
the ’317, ’010, and/or ’686 patents, in order to find inducement of
infringement. That is, you must find QUALCOMM is aware of the patent,
knows or should have known that the encouraged acts constitute
infringement of the patent, and has an intent to cause the encouraged
acts.

J.A. at 2194-95 (emphases added). This instruction is entirely consistent with our

guidance in DSU. Qualcomm objects, however, to the instruction that followed it:

2008-1199, -1271, -1272 23
 When considering whether Qualcomm knew of should have known that
the induced actions would constitute infringement, in the totality of the
circumstances, you may consider all of the circumstances, including
whether or not Qualcomm obtained the advice of a competent lawyer. I
will explain the significance of advice of counsel in more detail in a
moment.

Id. at 2195 (emphasis added). Following this instruction, the district court instructed the

jury as to contributory infringement, and then, in the context of instructions pertaining to

willfulness, the district court instructed the jury as follows:

In considering whether QUALCOMM acted in good faith, you should
consider all the circumstances, including whether or not QUALCOMM
obtained and followed the advice of a competent lawyer with regard to
infringement. The absence of a lawyer’s opinion, by itself, is insufficient to
support a finding of willfulness, and you may not assume that merely
because a party did not obtain an opinion of counsel, the opinion would
have been unfavorable. However, you may consider whether
QUALCOMM sought a legal opinion as one factor in assessing whether,
under the totality of the circumstances, any infringement by QUALCOMM
was willful.

Id. at 2199-2200 (emphasis added). This instruction comports with our holding in Knorr-

Bremse Systeme Fuer Nutzfahrzeuge, GmbH v. Dana Corp., where we held that there

is not “a legal duty upon a potential infringer to consult with counsel, such that failure to

do so will provide an inference or evidentiary presumption that such opinion would have

been negative.” 383 F.3d 1337, 1345 (Fed. Cir. 2004) (en banc).

Qualcomm argues, however, that the district court erred in allowing the

inducement verdicts to stand in light of its instruction to consider failure to obtain an

opinion of counsel as a factor in determining whether Qualcomm had the requisite level

of intent to induce infringement of Broadcom’s patents. Qualcomm’s argument

essentially rests on the proposition that Seagate altered the standard for establishing

the intent element of inducement. Qualcomm contends that because in Seagate we

2008-1199, -1271, -1272 24
“abandon[ed] the affirmative duty of due care” to avoid infringement and

“reemphasize[d] that there is no affirmative obligation to obtain opinion of counsel,” 497

F.3d at 1371, and because specific intent is a stricter standard than the “objective

recklessness” standard adopted in Seagate, evidence not probative of willful

infringement cannot be probative of specific intent to induce infringement. That is,

Qualcomm argues that opinion-of-counsel evidence is no longer relevant in determining

the intent of an alleged infringer in the inducement context. We disagree.

Although Qualcomm is correct that there is no affirmative duty to seek opinion of

counsel regarding infringement, and that it is improper to allow an “adverse inference or

evidentiary presumption that such an opinion would have been unfavorable,” Knorr-

Bremse, 383 F.3d at 1346, it is incorrect in arguing that Seagate altered the state of

mind requirement for inducement. Our en banc holding in DSU remains the relevant

authority on that point. Despite Qualcomm’s assertion that the intent standard for

inducement is higher than that for willful infringement, a lack of culpability for willful

infringement does not compel a finding of non-infringement under an inducement

theory. Qualcomm’s argument reflects a misunderstanding of our holding in DSU.

While inducement “requires more than just intent to cause the acts that produce

direct infringement,” and also requires “that the alleged infringer knowingly induced

infringement and possessed specific intent to encourage another’s infringement,” DSU,

471 F.3d at 1306, this intent may be established through circumstantial evidence, see

id. Moreover, “[t]he requisite intent to induce infringement may be inferred from all of

the circumstances.” Water Techs. Corp. v. Calco, Ltd., 850 F.2d 660, 669 (Fed. Cir.

1988). Additionally, we noted in DSU that this intent may be established where an

2008-1199, -1271, -1272 25
alleged infringer who “knew or should have known his actions would induce actual

infringements,” is shown to have induced infringing acts through his actions. 471 F.3d

at 1306 (emphasis added) (adopting the same two-part test articulated in Manville, 917

F.2d at 553). That is, the “affirmative intent to cause direct infringement,” id., required

by DSU may be shown—just as the jury was instructed in DSU itself—by establishing

first that the defendant “intended to cause the acts that constitute the direct

infringement,” and second that the defendant “kn[ew] or should have known [that] its

action would cause the direct infringement,” DSU, 471 F.3d at 1305. Because opinion-

of-counsel evidence, along with other factors, may reflect whether the accused infringer

“knew or should have known” that its actions would cause another to directly infringe,

we hold that such evidence remains relevant to the second prong of the intent analysis.

Moreover, we disagree with Qualcomm’s argument and further hold that the failure to

procure such an opinion may be probative of intent in this context. It would be

manifestly unfair to allow opinion-of-counsel evidence to serve an exculpatory function,

as was the case in DSU itself, see 471 F.3d at 1307, and yet not permit patentees to

identify failures to procure such advice as circumstantial evidence of intent to infringe.

Accordingly, we find no legal error in the district court’s jury instructions as they relate to

inducement.

Thus, the district court did not err in instructing the jury to consider “all of the

circumstances,” nor in instructing the jury to consider—as one factor—whether

Qualcomm sought the advice of counsel as to non-infringement. Even if the instructions

had been legally incorrect, Qualcomm failed to suggest any alternative instructions. To

the extent that Qualcomm argues that the district court should have stressed other

2008-1199, -1271, -1272 26
factors relevant to Qualcomm’s intent to induce infringement, it should have suggested

such instructions to the district court. See NTP, Inc. v. Research In Motion, Ltd., 418

F.3d 1282, 1311-12 (Fed. Cir. 2005) (noting that altering a jury verdict based on

erroneous jury instructions requires a party to demonstrate, inter alia, that “it requested

alternative instructions that would have remedied the error”).

Furthermore, we agree with Broadcom that substantial evidence supports the

jury’s verdict of induced infringement. Qualcomm seems to argue that the jury’s verdict

cannot stand because Broadcom failed to produce direct evidence of intent to induce

infringement. That is not the law. “A patentee may prove direct infringement or

inducement of infringement by either direct or circumstantial evidence. There is no

requirement that direct evidence be introduced, nor is a jury’s preference for

circumstantial evidence over direct evidence unreasonable per se.” Liquid Dynamics

Corp. v. Vaughan Co., 449 F.3d 1209, 1219 (Fed. Cir. 2006) (internal citation omitted).

Moreover, “[t]he drawing of inferences, particularly in respect of an intent-implicating

question . . . is peculiarly within the province of the fact finder that observed the

witnesses.” Rolls-Royce Ltd. v. GTE Valeron Corp., 800 F.2d 1101, 1110 (Fed. Cir.

1986); see also Fuji Photo Film Co. v. Jazz Photo Corp., 394 F.3d 1368, 1378 (Fed. Cir.

2005) (declining to disturb jury’s verdict because intent to induce infringement “is a

factual determination particularly within the province of the trier of fact”).

Qualcomm does not dispute that it was on notice of Broadcom’s patents and

infringement contentions. And it concedes that it “worked closely with its customers to

develop and support the accused products, and that Qualcomm did not make changes

to those products or give its customers [instructions regarding how to avoid

2008-1199, -1271, -1272 27
infringement] after this lawsuit was filed.” Appellant’s Supplemental Br. at 20-21.

Qualcomm also fails to identify evidence in the record to counter the district court’s

related observations pertaining to willfulness: “[T]he totality of the circumstances

presented in the evidence supports the jury’s findings: a failure to investigate, a failure

to explore design around approaches, a failure to take remedial steps—and, of course,

a failure to seek legal advice. This was sufficient for the jury to infer a lack of good

faith.” JMOL Opinion at 8. Although Qualcomm stresses that it did obtain opinions of

counsel regarding invalidity of the patents in suit, it concedes that the district court

properly excluded this fact from evidence in light of Qualcomm’s decision not to waive

privilege with respect to these opinions.

Taken as a whole, this record provides substantial evidence to support the jury’s

verdict and is consistent with our precedent. See, e.g., Golden Blount, Inc. v. Robert H.

Peterson Co., 438 F.3d 1354, 1364 n.4 (Fed. Cir. 2006) (noting that it was “undisputed

that Peterson had notice of the patent . . . and that Peterson provided [an] instruction

sheet to customers directing them to perform specific acts leading to the assembly of

infringing devices, from which the district court could draw an inference of intent”);

Mentor H/S, Inc. v. Med. Device Alliance, Inc., 244 F.3d 1365, 1379 (Fed. Cir. 2001)

(upholding a jury verdict of induced infringement where the defendant was aware of the

patent yet chose to continue selling products designed specifically for use in an

infringing manner, where it was apparent that the jury discredited the defendant’s

assertion that it relied upon a non-infringement opinion letter in good faith). Thus, we

will not disturb the jury’s inference that Qualcomm possessed the specific intent

necessary to induce the infringement of Broadcom’s patents.

2008-1199, -1271, -1272 28
 Because we do not agree that a new trial on inducement is warranted, we need

not consider Qualcomm’s arguments regarding liability and damages issues allegedly

interwoven with the inducement claim.

E. Permanent Injunction

Finally, Qualcomm argues that Broadcom is not entitled to the permanent

injunction issued by the district court. Qualcomm asks us to vacate the injunction

against Qualcomm CDMA2000 products based on the four-factor test outlined in eBay

Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006). Qualcomm argues that because

Broadcom makes WCDMA chips, not CDMA2000 products like Qualcomm, it could not

have established irreparable injury from Qualcomm’s sale of CDMA2000 chips. Similar

arguments are made with respect to the QChat software, because Broadcom does not

offer a PTT product. Qualcomm contends that Broadcom’s license agreement with

Verizon demonstrates that monetary remedies may adequately compensate it for

infringement injuries. It further argues that the district court clearly erred in balancing

the hardships because the injunction will disrupt Qualcomm’s ongoing business model

with CDMA carriers and because the sunset provisions do not allow sufficient time to

develop design-around technology. Finally, Qualcomm contends that the public interest

factor should weigh in its favor because of the risk to downstream carriers that rely on

Qualcomm’s CDMA2000 technology, especially carriers like Sprint that will suffer

significant disadvantages as compared to Verizon, which has a license with Broadcom.

An amicus brief filed by Sprint presents similar arguments, and primarily contends that

the district court should have considered an ongoing royalty in product areas in which

Broadcom does not participate, such as PTT technology and CDMA2000 chips in

general. See Paice LLC v. Toyota Motor Corp., 504 F.3d 1293, 1314 (Fed. Cir. 2007)

2008-1199, -1271, -1272 29
(noting that “[u]nder some circumstances, awarding an ongoing royalty for patent

infringement in lieu of an injunction may be appropriate”).

Broadcom responds that the district court took the factors identified by

Qualcomm into account and found persuasive several public statements by Qualcomm

and its downstream customers indicating that design-arounds were in the works,

suggesting that the impact of the injunction was not critical. Broadcom argues that—

irrespective of whether it sells identical products or chipsets under the same protocols—

it competes indirectly with Qualcomm based on the nature of the market, and that

Qualcomm has admitted that fact in other contexts. It contends that its business model

of generally not licensing its patent portfolio favors injunctive relief, and that its license

with Verizon was designed to ameliorate the impact of an injunction and in any case

cannot be compared to a license with a direct competitor like Qualcomm. It also argues

that sunset provisions’ allowance of infringement for up to twenty months after the jury

verdict adequately addresses concerns related to the balance of hardships and public

interest.

“[W]e review a district court’s decision to grant an injunction and the scope of

that injunction for abuse of discretion.” Joy Techs., Inc. v. Flakt, Inc., 6 F.3d 770, 772

(Fed. Cir. 1993).

According to well-established principles of equity, a plaintiff seeking a
permanent injunction must satisfy a four-factor test before a court may
grant such relief. A plaintiff must demonstrate: (1) that it has suffered an
irreparable injury; (2) that remedies available at law, such as monetary
damages, are inadequate to compensate for that injury; (3) that,
considering the balance of hardships between the plaintiff and defendant,
a remedy in equity is warranted; and (4) that the public interest would not
be disserved by a permanent injunction.

2008-1199, -1271, -1272 30
eBay, 547 U.S. at 391. Here, the district court provided a well-reasoned and

comprehensive opinion addressing injunctive relief, in which it accounted for the unique

issues presented by the various patents and the infringing products and each of the

factors re-affirmed in eBay. We address each factor in turn.

1. Irreparable Injury

Qualcomm argues that because Broadcom does not sell or plan to sell

CDMA2000 chips, it cannot allege harm resulting from Qualcomm’s CDMA200 chip

sales and that Broadcom’s arguments on this point amount to no more than speculative,

unsubstantiated assertions of harm. Similarly, Qualcomm argues that Broadcom’s lack

of a PTT product precludes a finding of irreparable injury stemming from the sale of

products featuring the QChat software. Qualcomm also argues that Broadcom’s license

to Verizon—a CDMA carrier—undermines its claim of irreparable injury.

Broadcom argues that Qualcomm itself has admitted that it competes indirectly

with Broadcom, despite offering different technology in its chipsets. Broadcom also

contends that the district court correctly considered Broadcom’s general policy of not

licensing its patents and the harm that would ensue from a compulsory license to its

most significant competitor. Broadcom further argues that it entered into a license

agreement with Verizon, in part, to minimize the potential impact of an injunction to third

parties or consumers while Qualcomm designs around Broadcom’s patents.

The district court found that this factor weighed in favor of an injunction with

respect to both the ’317 and ’010 patents. Injunction Opinion at 10, 15. It remains an

open question “whether there remains a rebuttable presumption of irreparable harm

following eBay,” Amado v. Microsoft Corp., 517 F.3d 1353, 1359 n.1 (Fed. Cir. 2008),

2008-1199, -1271, -1272 31
but the district court did not abuse its discretion in finding irreparable injury here even if

Broadcom benefits from no such presumption. The district court explained that the

evidence at trial demonstrated that

[t]he market for baseband chips is unlike the typical market for consumer
goods where competitors compete for each consumer sale, and the
competition is instantaneous and on-going. . . . Competition for sales is
not on a unit-by-unit basis, but rather competition is characterized by
competing for “design wins” for the development and production of cell
phones which will embody the proposed chip. . . . In this kind of a market,
the exclusion has a competitive effect on a firm even if it does not have an
immediately available product.

Injunction Opinion at 5-6. And as identified by Broadcom, Qualcomm has previously

conceded this indirect competition. E.g., J.A. at 7785 n.7 (“Any cell phone customer

who has chosen between, e.g., Cingular (GSM) and Verizon (CDMA) service knows

that handsets and thus chipsets do compete and thus ‘substitute’ across standards.”).

Thus, Broadcom provided evidence of irreparable harm, despite the fact that it

does not currently practice the claimed inventions. This result is consistent with eBay,

in which the Supreme Court cautioned that “traditional equitable principles do not permit

such broad classifications” as presuming that a patentee cannot establish irreparable

harm based on a patentee’s “willingness to license its patents” or “its lack of commercial

activity in practicing the patents.” 547 U.S. at 393; see also Rite-Hite Corp. v. Kelley

Co., 56 F.3d 1538, 1547 (Fed. Cir. 1995) (en banc) (noting that “[t]here is no

requirement in this country that a patentee make, use, or sell its patented invention”

(citing Cont’l Paper Bag Co. v. E. Paper Bag Co., 210 U.S. 405, 424-30 (1908))).

Although we recently upheld a district court’s denial of a permanent injunction where the

patentee “had attempted to prove irreparable injury by alleging irreparable harm to his

exclusive licensee, rather than himself,” we noted that “patent owners that license their

2008-1199, -1271, -1272 32
patents rather than practice them ‘may be able to satisfy the traditional four-factor test’

for a permanent injunction.” Voda v. Cordis Corp., 536 F.3d 1311, 1329 (Fed. Cir.

2008) (first emphasis added) (quoting eBay, 547 U.S. at 393). Reliance on Voda is

misplaced in this case because Broadcom does not rely on harm to others, but rather

alleges that its own commercial activities will be irreparably injured by Qualcomm’s

infringing activity. As such, the district court did not abuse its discretion in finding that

allowing Qualcomm to sell CMDA2000 chips implementing Broadcom’s patented

features would harm Broadcom’s efforts to market WCDMA solutions. See Injunction

Opinion at 15.

2. Lack of Adequate Remedy at Law

As with the irreparable injury prong of the analysis, the district court found that

this prong weighed in favor of an injunction with respect to all of the infringed patents.

Id. at 11, 17. Qualcomm relies primarily on the Verizon license agreement as evidence

that Broadcom can be adequately compensated by monetary damages. Broadcom

responds that the Verizon agreement provided many non-monetary benefits to

Broadcom, such as the formation of a strategic business alliance, and that a license to

Qualcomm’s customer is of limited probative value as to its willingness to enter into a

similar such agreement with its direct competitor.

We agree with Broadcom that the district court did not abuse its discretion here,

particularly in light of its findings that “the structural nature of a ‘design win’ market

favors a finding that monetary damages are inadequate.” Id. at 16. We also agree that

the Verizon license has little bearing on the effect of a compulsory license to a direct

competitor, particularly in light of these market realities. Additionally, while Broadcom’s

2008-1199, -1271, -1272 33
damages expert testified that any comparison of Qualcomm’s lost profits based on an

injunction to Broadcom’s potential gain would be “almost pure speculation at this point,”

J.A. at 3217, this difficulty in estimating monetary damages reinforces the inadequacy of

a remedy at law. Cf. Jacobsen v. Katzer, 2008 WL 3395772, at *6 (Fed. Cir. Aug 13,

2008) (noting, in the copyright license context, that “because a calculation of damages

is inherently speculative, these types of license restrictions might well be rendered

meaningless absent the ability to enforce through injunctive relief”).

3. Balance of Hardships

The district court acknowledged the harm inflicted on Qualcomm through a

permanent injunction, but held that “with a sunset provision which ameliorates the

negative effects on Qualcomm, the balance of hardships favors Broadcom” with respect

to the ’317 patent. Injunction Opinion at 16. It found this factor to be neutral regarding

the ’010 patent. Id. at 10. At the outset, we note that “[o]ne who elects to build a

business on a product found to infringe cannot be heard to complain if an injunction

against continuing infringement destroys the business so elected.” Windsurfing Int’l,

Inc. v. AMF, Inc., 782 F.2d 995, 1003 n.12 (Fed. Cir. 1986). We agree with the district

court that Qualcomm should not be permitted to prevail on a theory “that successful

exploitation of infringing technology shields a party from injunctive relief,” Injunction

Opinion at 16, but in any event we find this factor to be neutral at best in light of the

district court’s carefully constructed sunset provisions. The district court provided for a

twenty-month sunset provision from the date of the May 2007 jury verdict. See id. at 12,

18. This does not amount to an abuse of discretion, particularly in light of the district

court’s finding that “the time line from a ‘win’ to actually bringing the product to the

2008-1199, -1271, -1272 34
consumer is about 18 months,” id. at 5, and in light of the fact that Qualcomm could

have—and in fact may have—begun design-around efforts upon receipt of the

complaint, which Broadcom filed more than two years prior to the jury verdict.

4. Public Interest

The district court found that although it is generally in the public interest to uphold

patent rights, see generally Rite-Hite, 56 F.3d at 1547, “an immediate permanent

injunction would adversely affect the public” with respect to the ’010 patent, Injunction

Opinion at 9, and that “an immediate permanent injunction would adversely affect

network carriers and handset manufacturers that currently employ chips which infringe

the ’317 Patent in their products,” id. at 17. However, the district court held that the

aforementioned sunset provisions “balance[] the policy of protecting the patentee’s

rights against the desirability of avoiding immediate market disruptions.” Id. at 18. We

agree that the sunset provisions mitigate the harm to the public and that the district

court did not abuse its discretion in fashioning a remedy that protects Broadcom’s rights

while allowing Qualcomm time to develop non-infringing substitutes. See Verizon

Servs. Corp. v. Vonage Holdings Corp., 503 F.3d 1295, 1311 n.12 (Fed. Cir. 2007)

(noting that a sunset provision would have allowed time for an infringer to implement

non-infringing substitutes).

* * *

In sum, the district court did not abuse its discretion in issuing the permanent

injunction against Qualcomm. However, we reverse the injunction and damages award

as they pertain to the ’686 patent and remand to the district court so that the injunction

and damages may be adjusted in light of our reversal of the jury verdict of infringement.

2008-1199, -1271, -1272 35
 III. CONCLUSION

For the above reasons, we conclude that the district court erred in construing

claim 3 of the ’686 patent, which we hold is invalid under the proper construction; that

substantial evidence supports the jury’s findings of infringement and validity as to the

’317 and ’010 patents; and that the district court did not abuse its discretion in issuing a

permanent injunction. Thus, its judgment is

AFFIRMED-IN-PART, REVERSED-IN-PART, and REMANDED.

2008-1199, -1271, -1272 36