Wang Laboratories, Inc. v. Mitsubishi Electronics America, Inc. And Mitsubishi Electric Corporation, Defendants/cross-Appellants

RICH, Circuit Judge.

Appellant Wang Laboratories, Inc. sued Mitsubishi Electronics America, Inc. for infringement of two patents. Mitsubishi and its parent, Mitsubishi Electric Corporation, (collectively “Mitsubishi”) filed a declaratory judgment action and the cases were consolidated in the United States District Court for the Central District of California. The district court granted Mitsubishi partial summary judgment of non-infringement of one patent. A jury then found both patents not invalid and that Mitsubishi literally infringed the other patent. But the jury concluded that Wang’s conduct in dealing with Mitsubishi created an implied license from Wang to Mitsubishi to practice the invention claimed in the patent held infringed. Wang appeals and Mitsubishi cross-appeals. Since we agree that Mitsubishi was licensed, there was no infringement and we therefore affirm.

BACKGROUND

The two patents in suit relate to memory modules known as “Single In-line Memory Modules” or SIMMs. Generally, Wang’s SIMM design consists of nine memory chips with one reserved for parity checking. The module is sometimes described as a “times nine” or “X9” module. The parity chip independently stores additional information allowing a device to verify the validity of data stored in and retrieved from the other eight memory chips. The nine chips are encased in plastic carriers having leads (“leaded”) and are arranged in one row, along with associated decoupling capacitors to protect the chips, on a substrate with 30 connective terminal pins. When mounted on the printed-circuit “mother board” of a computer containing the computer’s central processing unit, for example, these modules allow the computer to write and read data in binary format to and from the SIMMs for storage. In developing its SIMM, Wang used 64-kilobit (“64K”) memory chips. Together, eight 64K memory chips could store 512 kilobits or 64 kilobytes of data (eight binary digits or “bits” constitute one “byte”). More recent SIMMs may contain 256-kilobit (“256K”) or larger memory chips.

James Clayton, the named inventor in the patents in suit, joined Wang Laboratories in the fall of 1982. At the time, computer memory components remained relatively large, expensive, and difficult to upgrade. In the spring of 1983, Clayton developed the SIMM as a smaller, lower cost, replaceable form of computer memory. On September 2, 1983, Clayton, with Wang as assignee, applied for a patent on the SIMM invention, application serial number 528,817. The original application contained three claims, all of which the Patent and Trademark Office (“PTO”) rejected on June 13, 1985. The PTO based its rejection, in part, on obviousness of the claimed subject matter. Wang amended the application to cancel claims 1, 2, and 3, and to add claims 4, 5, and 6. The new claims clarified that the invention consisted of a single row of memory chips packaged in “plastic leaded chip carriers” for mounting on “epoxy-glass” printed circuit boards. The PTO’s final office action, however, rejected the amended claims, this time solely based on obviousness.

Wang submitted a file wrapper continuation application, serial number 873,879, on June 12, 1986, replacing claims 4, 5, and 6 with claim 7. Wang then abandoned the initial application. In “remarks” accompanying the continuation application, Wang argued that the single new claim recited a unique, specific combination of elements distinguishable from the prior art, in particular from “[t]he EDI EDH-4816 publication describing] a single in-line memory module *1574which consists of eight ceramic chips mounted on a ceramic substrate. The module contains thirty-two terminal pins and ten decoupling capacitors on the main voltages.” In contrast, Wang claimed “nine memory chips (eight for data, one for parity) contained in plastic leaded chip carriers and mounted on a[sie] epoxy-glass pc [printed circuit] board of particular dimensions. The module has thirty terminal pins and eight decoupling capacitors connected between the memory chips to suppress voltage spikes.” The application resulted in patent No. 4,656,605 (’605 patent), issued April 7, 1987. The single claim reads (emphasis ours):

1. A memory module for installation on a printed circuit motherboard comprising:
eight data memory chips for storing digital data, each having a data input and output, a control input, and an address input, and each being packaged in a plastic leaded chip carrier;
a ninth memory chip for storing error detection and correction information associated with the eight data memory chips, said ninth memory chip having a data input and output, a control input and an address input interconnected with those of the eight memory chips, and a control input to provide writing in or reading out of the ninth memory chip at times other than when said bytes of digital information are written into or read out of the eight data memory chips to thereby facilitate said error detection and correction operation;
an epoxy-glass printed circuit board substrate having a length and width adequate for mounting thereon only in a single row said nine memory chips and for interconnecting the control inputs and the address inputs of the memory chips so that bytes of digital information may be input to or output from the memory chips one at a time;
the substrate including thirty terminals for providing access to the data inputs and outputs, control inputs, and address inputs of the nine memory chips to enable reading and writing of bytes of digital information into and out of the eight memory chips and to enable reading and writing of error detection and correction information into and out of the eight memory chips;
support means for supporting the memory module at an angle with respect to the printed circuit motherboard when the memory module is installed thereon; and
eight decoupling capacitors, mounted on said substrate and connected between the nine memory chips, for suppressing transient voltage spikes between said memory chips.

Wang filed another application, on February 20, 1987, as a continuation of application number 873,879. This application comprised two claims and led to patent No. 4,727,513 (’513 patent), which issued on February 23, 1988. The two claims of the ’513 patent are as follows:

1. A memory module for installation on a printed circuit motherboard comprising
nine data memory chips for storing digital data, each having a data input and output, control input, and an address input, and each being packaged in a plastic leaded chip carrier, wherein said ninth memory chip is for storing detection and correction information associated with the eight data memory chips,
an epoxy-glass printed circuit board substrate having a length and width adequate for mounting thereon only in a single row said nine memory chips and for interconnecting the control inputs and the address inputs of the memory chips so that bytes of digital information may be input to or output from the memory chips,
the substrate including thirty terminals for providing access to the data inputs and outputs, control inputs, and address inputs of the nine memory chips and to enable reading and writing of information into and out of the nine chips,
support means for supporting the memory module at an angle with respect to a motherboard and
decoupling capacitors mounted on said substrate and coupled to the memory chips for suppressing transient voltages.
2. The module of claim 1 wherein all nine memory chips are interconnected such that data is input to or output from the ninth *1575memory chips [sic] when data is input to or output from the other eight memory chips,

A panel of Wang employees, including Clayton, introduced the SIMM technology to members of the computer industry press in June 1983. After giving a detailed description of their invention, Wang personnel answered questions. According to the panelists, Wang hoped to preempt the anticipated future introduction of 256K chips by Japanese manufacturers in a format known as the Small Outline Integrated Circuit or SOIC. By combining eight existing 64K chips into a 64-kilobyte package, Wang could introduce immediately a smaller, denser, lower cost memory option. The panel went on to note an expected struggle at the Joint Electronic Device Council (“JEDEC”), a committee of the Electronic Industries Association invested with responsibility for setting industry standards, over which format would become the memory module standard.

Clayton described the period preceding the upcoming JEDEC meetings, with manufacturers awaiting indication of which way the market would go before committing to a module design. He explained Wang’s belief that the present introduction of its new package should be the catalyst to sway the industry to its approach. Clayton revealed that Wang would not produce SIMMs. Instead, Wang intended to encourage others to produce the modules and would then buy SIMMs for use in its products. The panel indicated that companies already were preparing to make SIMMs. In response to questions, a panelist stated that Wang was not seeking patent rights in the SIMM,1 that no licensing agreements were involved for the companies approached by Wang to make SIMMs, and that SIMM makers could sell their products to third parties. Panelist Daniel Devlin summarized Wang’s goal: “Hopefully if they sell it, and if they have enough interest, and the market gets big enough, we get the advantage of the cost reduction because of the volumes involved.” Numerous trade publications carried SIMM stories in their July or August 1983 issues.

Wang subsequently brought its SIMMs to JEDEC and sought to persuade the group to designate Wang’s design a standard. Wang argued for adoption of the SIMM from September 1983 through June 1986, when JE-DEC accepted the SIMM as a standard. During that period, Wang did not inform JEDEC of its ongoing pursuit of patent rights in the SIMM. Meanwhile, several manufacturers cooperated with Wang to begin mass producing and marketing SIMMs. As predicted in the 1983 press presentation, a large market developed for the modules and Wang became a high volume purchaser.

Mitsubishi first met with Wang regarding SIMMs in December 1983. In their meetings, Wang supplied drawings and other details to Mitsubishi and repeatedly requested that Mitsubishi manufacture SIMMs. Mitsubishi researched the possibility of producing SIMMs for Wang, but did not proceed with the project at that time. After Mitsubishi began making 256K memory chips, however, Mitsubishi decided to assemble 256K SIMMs, incorporating, the new chips into a SIMM similar to Wang’s, for sale to Wang and others. Mitsubishi declined to assess engineering costs to Wang, contrary to its asserted practice when creating a custom product exclusively for a particular purchaser.

In 1985 meetings, in the context of ongoing contacts between the two companies, Mitsubishi and Wang discussed Mitsubishi’s new 256K SIMMs.2 In one meeting, Clayton suggested that Mitsubishi modify its SIMM by placing the decoupling capacitors on the same side of the substrate with the chips, as in the original Wang design. Mitsubishi complied. Mitsubishi went on to mass produce 256K SIMMs; and in 1987, Wang began buying Mitsubishi SIMMs. Wang never informed Mitsubishi of its patent applications, *1576patents, or of any intent to execute a license or receive royalties until a December 22, 1989 letter accusing Mitsubishi of infringing the ’605 and ’513 patents, which had issued in 1987 and 1988 respectively.

Wang sued Mitsubishi for infringement on June 4, 1992, in the United States District Court for the Eastern District of Virginia. Wang argued that Mitsubishi literally infringed the ’513 patent and infringed the ’605 patent under the doctrine of equivalents. Mitsubishi had the case transferred to the United States District Court for the Central District of California and filed counterclaims seeking a declaratory judgment of invalidity, non-infringement, and unenforceability, as well- as alleging assorted state and federal antitrust violations. Mitsubishi also asserted several affirmative defenses, including the defense that Wang’s conduct created an implied license.

Wang moved for summary judgment that Mitsubishi infringed the ’605 patent under the doctrine of equivalents because the nine decoupling capacitors beneath nine memory chips in Mitsubishi’s SIMMs are equivalent to the “eight decoupling capacitors ... between the nine memory chips” claimed in Wang’s ’605 patent. The court denied Wang’s motion. Mitsubishi then sought summary adjudication that prosecution history estoppel barred Wang .from relying on the doctrine of equivalents with regard to this “eight capacitor limitation.” For support, Mitsubishi focused on Wang’s June 12, 1986, amendment and accompanying remarks distinguishing claim 7 from the EDI EDH-4816 publication disclosing 10 capacitors.

The district court agreed that Wang relied on the “unique combination of elements” in its design, including exactly eight capacitors, and granted partial summary judgment of non-infringement by Mitsubishi of the ’605 patent.

The parties tried the remaining issues to a jury. At the close of evidence, each side was denied judgment as a matter of law (“JMOL”) under Rule 50(a). Next, Wang and Mitsubishi fought over the formulation of the implied license instruction to the jury. Wang submitted legal estoppel language; Mitsubishi objected and offered equitable estoppel or estoppel by conduct as the basis of its license. The final instruction merged the parties’ competing proposals into one instruction which read as follows:

One who owns a patent as patentee or assignee, having the right to exclude others from making, using, or selling what is claimed-, may agree to let another do one or more of those acts. This is called a license, and the person allowed to do the set of acts is a licensee.
An implied license is a form of implied-in-fact contract. In order to proved [sic] the defense of implied license, [Mitsubishi] must establish by a preponderance of the evidence that (1) there was an existing relationship between Wang and [Mitsubishi]; (2) within that relationship, Wang transferred a right to use SIMM invention to [Mitsubishi]; (3) the right was transferred for valuable consideration; and (4) Wang has now denied the existence of the right it transferred to [Mitsubishi].
In deciding whether [Mitsubishi has] proved the existence of an implied license, you may consider statements and conduct of Wang from which one would reasonably infer Wang’s consent to Mitsubishi’s making, using, or selling products to persons other than Wang under the patents.
If you find that a right to make, use, or sell the SIMM invention was granted to [Mitsubishi] then you should find and [sic] implied license was granted to [Mitsubishi] by Wang.

The relevant question to the jury on the Special Verdict Form read, “Has Mitsubishi proven by a preponderance of the evidence that Wang licensed Mitsubishi to make, use, or sell the subject matter of the ’513 patent?” The jury answered “Yes.” On other issues, the jury determined that Mitsubishi failed to prove that the ’605 or ’513 patents, were invalid for obviousness or for the alleged nondisclosure of the inventor’s best mode. The jury further found that the Mitsubishi SIMMs infringed the ’513 patent claims, but that infringement was not willful. In- its advisory capacity, the jury concluded that Mitsubishi did not prove its defenses of inequitable conduct, equitable estoppel, laches, patent misuse, or unclean hands.

*1577After the verdict, the court held a hearing on Mitsubishi’s equitable defenses, then directed Wang to draft a proposed judgment and findings of fact and conclusions of law adopting the jury’s determinations on the equitable defenses, which Wang did. Mitsubishi objected to Wang’s drafts and protested that, although the findings and conclusions did not cover the implied license, Wang wrote findings on other issues which contradicted and undermined the jury’s implied license verdict. Furthermore, Wang had written that the jury found an implied license, “under the doctrine of legal estoppel.” Notwithstanding Mitsubishi’s objections, the court entered both the Judgment and the Findings of Fact and Conclusions of Law.

On March 6, 1995, the parties argued six motions to the district court. Wang’s motions addressed the implied license defense. First, Wang moved for Rule 50(b) JMOL because Wang believed that no reasonable jury could find Mitsubishi provided valuable consideration for a license, as required by Wang’s formulation of implied license elements based on legal estoppel. Second, Wang sought a partial new trial on the implied license defense. Third, Wang moved to amend the judgment based on its theory that the implied license conclusion mandated a finding that an implied contract existed and that, assuming an absence of other consideration, Mitsubishi must have promised royalties to Wang.

Mitsubishi moved for JMOL contending that because Clayton admitted he knew a best mode which comprised two power pins and two ground pins wired in a symmetrical configuration, and because the patent contains no express disclosure of that wiring scheme, the jury erred in finding Mitsubishi failed to prove that Wang concealed the inventor’s best mode. Next, Mitsubishi asserted that the evidence required a finding of obviousness and, therefore, invalidation of Wang’s patents under 35 U.S.C. § 103. Last, Mitsubishi argued that five differences between Mitsubishi’s modules and the SIMM claimed in the ’513 patent bar a finding of literal infringement of the ’513 patent. The court denied all six motions.

ANALYSIS

Wang appeals the partial summary judgment of non-infringement of the ’605 patent based on prosecution history estoppel, and the resolution of the implied license issue to negate infringement of the ’513 patent. Mitsubishi cross-appeals from the judgment on best mode, nonobviousness, and infringement.

The ’605 Patent: Prosecution History Estoppel

Wang asserts that the district court erred when it ruled that prosecution history estoppel limits the scope of the ’605 patent to SIMMs with exactly eight capacitors. Since the estoppel ruling prevents Wang from resorting to the doctrine of equivalents to extend the patent’s scope to cover nine capacitors, and given that Mitsubishi’s modules with nine capacitors do not literally infringe the ’605 patent, the court granted partial summary judgment providing that Mitsubishi’s modules with nine capacitors do not infringe the.’605 patent. The issue of whether the estoppel extends to the ’513 patent is not before us. See Mark I Mktg. Corp. v. R.R. Donnelley & Sons Co., 66 F.3d 285, 291, 36 USPQ2d 1095, 1100 (Fed.Cir.1995), cert. denied, — U.S. -, 116 S.Ct. 917, 133 L.Ed.2d 847 (1996) (extended estoppel from parent to patent from continuation application). Summary judgment is appropriate when no genuine issue of material fact exists and judgment is proper as a matter of law. Fed.R.Civ.P. 56(c). We review grants of summary judgment de novo. Conroy v. Reebok Int'l Ltd., 14 F.3d 1570, 1575, 29 USPQ2d 1373,1377 (Fed.Cir.1994).

Application of the doctrine of equivalents may allow a patentee to recover for infringement though the accused device falls outside the literal scope of the claims if the differences between the claimed invention and the device are insubstantial. See, e.g., Pall Corp. v. Micron Separations, Inc., 66 F.3d 1211, 1218, 36 USPQ2d 1225, 1229-30 (Fed.Cir.1995). Prosecution history estoppel acts as one check on application of the doctrine of equivalents, Mark I, 66 F.3d at 291, 36 USPQ2d at 1100, by precluding a patentee from regaining, through litigation, coverage *1578of subject matter relinquished during prosecution of the application for the patent. Zenith Lab., Inc. v. Bristol-Myers Squibb Co., 19 F.3d 1418, 1424, 30 USPQ2d 1285, 1290 (Fed.Cir.1994). Whether estoppel applies is a question of law. LaBounty Mfg., Inc. v. United States Int’l Trade Comm’n, 867 F.2d 1572, 1576, 9 USPQ2d 1995, 1998 (Fed.Cir. 1989). We examine the statements and actions of the patentee before the PTO during prosecution, Haynes International, Inc. v. Jessop Steel Co., 8 F.3d 1573, 1578, 28 USPQ2d 1652, 1656 (Fed.Cir.1993); Pall, 66 F.3d at 1219, 36 USPQ2d at 1230, and ask what a competitor reasonably may conclude the patentee surrendered to gain issuance of the patent. Haynes, 8 F.3d at 1578, 28 USPQ2d at 1656. Arguments and amendments made to secure allowance of a claim, especially those distinguishing prior art, presumably give rise to prosecution" history estoppel. Pall, 66 F.3d at 1219, 36 USPQ2d at 1230. Once prosecution history estoppel limits the scope of a patent, the patentee may not recover for infringement where infringement would require an equivalence between a claim element and an aspect of the accused item that falls within the estoppel. South-wall Technologies, Inc. v. Cardinal IG Co., 54 F.3d 1570, 1579, 34 USPQ2d 1673, 1679-80 (Fed.Cir.), cert. denied, — U.S. -, 116 S.Ct. 515, 133 L.Ed.2d 424 (1995).

Wang amended its claims to distinguish its SIMMs from prior art. During the prosecution leading to issuance of the ’605 patent, Wang twice faced rejection of its claims for obviousness. Following the final office action rejecting all three claims and citing the EDI EDH-4816 publication, Wang submitted a file wrapper continuation application that condensed the definition of the invention into a single claim. The new claim added several limitations including, for the first time, exact numbers of terminals and decoupling capacitors. In remarks accompanying the new claim, Wang identified the 30 terminals and the eight decoupling capacitors as features distinguishing its claimed SIMM from the memory module disclosed in the EDI publication with 32 terminals and 10 decoupling capacitors. Wang emphasized the “unique combination of elements” in its SIMM, and cited extensive “exact copying of the invention” as evidence of patentability. After considering Wang’s submission, the PTO allowed the narrowed claim.

Wang’s actions before the PTO trigger prosecution history estoppel. The PTO rejected claims as obvious in light of prior art; the patentee amended, adding limitations arguing distinctions from the prior art which persuaded the PTO to allow the amended claim. In Wang Laboratories, Inc. v. Toshiba Corp., 993 F.2d 858, 26 USPQ2d 1767 (Fed.Cir.1993), for example, an earlier lawsuit for infringement of the same two patents, this court held that Wang had limited the scope of its patents to SIMMs with (1) exactly nine memory chips (2) in a single row. Id. at 868, 993 F.2d 858, 26 USPQ2d at 1776. We held that prosecution history estoppel attached because Wang had claimed its invention with precision to distinguish a plurality of prior art references in the field of this invention. Id. Similarly, we now hold that Wang likewise limited its claim to exactly eight capacitors. A competitor such as Mitsubishi would reasonably conclude that Wang surrendered coverage of SIMMs not meeting the eight capacitors limitation. Therefore, we affirm the district court’s grant of partial summary judgment of non-infringement of the ’605 patent.

The ’513 Patent: Implied License

Mitsubishi argued at trial that Wang’s conduct had created an implied license to it under the ’513 patent. The parties agreed to submit Mitsubishi’s implied license defense to the jury and the jury found an implied license to exist. The district court entered judgment on the verdict, and added the rationale that the implied license arose “under the doctrine of legal estoppel.” Focusing on the legal estoppel rationale, Wang moved for JMOL or a new trial on the implied license issue because, according to Wang, a reasonable jury could not have found facts to support legal estoppel and because the court misapplied the law of legal estoppel. See Fed.R.Civ.P. 50(b) and 59(a). Wang also moved the court to infer contract terms requiring Mitsubishi to pay royalties to Wang as part of the implied license. See Fed.R.Civ.P. 59(e). The district *1579court denied the motions. Wang now seeks a reversal of the judgment or new trial on the implied license defense. In an appeal from a judgment following denial of JMOL, we must affirm factual findings that are supported by substantial evidence, we then review legal conclusions to determine whether the factual findings can bear them. Railroad Dynamics, Inc. v. A. Stucki Co., 727 F.2d 1506, 1512, 220 USPQ 929, 935-36 (Fed.Cir.1984); Perkin-Elmer Corp. v. Computervision Corp., 732 F.2d 888, 893, 221 USPQ 669, 672-73 (Fed.Cir.1984). We review the denial of a motion for a new trial for abuse of discretion. Railroad Dynamics, 727 F.2d at 1512, 220 USPQ at 935-36.

Because the second step of this review will require us to determine whether the factual findings support the conclusion of law, we must first ascertain what facts the jury found relating to Mitsubishi’s implied license defense. Id. The Special Verdict Form asked the jury to answer only the ultimate question of whether an implied license exists, so the jury instructions serve as our most direct guide to what the jury decided.3 See Newell Cos., Inc. v. Kenney Mfg. Co., 864 F.2d 757, 765, 9 USPQ2d 1417, 1423 (Fed.Cir.1988) (we accept factual findings presumed from jury verdict, subject to substantial evidence test). In order to arrive at its affirmative answer on the implied license question, the jury necessarily found that (1) a relationship existed between Wang and Mitsubishi, (2) within that relationship, Wang granted to Mitsubishi a right to use its SIMM inventions, (3) Wang received valuable consideration for that grant of right, (4) Wang denied that Mitsubishi had an implied license, and (5) Wang’s statements and conduct created the impression that Wang consented to Mitsubishi making, using, or selling Wang’s patented inventions, including sales to consumers other than Wang. Separately, the jury found that Mitsubishi made, used, or sold several versions of SIMMs that would infringe the ’513 patent but for the license.

With regard to the factual component of our review, Wang challenges only the finding of valuable consideration. Before we can apply the substantial evidence test to this finding, however, it becomes necessary to look beyond the jury instructions and verdict to discover what the jury understood to constitute “valuable consideration.” The district court’s order denying Wang’s motion for JMOL identified three benefits conferred on Wang: (1) by agreeing to manufacture and sell SIMMs, Mitsubishi .contributed to a “high volume supply and downward pressure on [the] price” of SIMMs, which benefited Wang as a purchaser of SIMMs; (2) Mitsubishi absorbed development and tooling costs; and, (3) Mitsubishi redesigned its SIMMs to conform to Wang’s preferred design. The district court ruled that each form of consideration was supported by substantial evidence.

Looking to the record in search of substantial evidence, Wang’s own statements show that when the company introduced its SIMM design in June 1983, Wang intended to buy SIMMs from other producers rather than produce SIMMs itself. Eventually, Wang hoped to “get the advantage of the cost reduction because of the volumes involved.” The evidence indicates that Wang considered lower prices and a larger market to be of value to it, as they obviously were as a large user of SIMMs. In time, JEDEC adopted Wang’s SIMM design as a standard in what became a multi-billion dollar market. The market grew; prices dropped. The record contains evidence that Mitsubishi contributed to this advantageous outcome. Clayton’s notes about Mitsubishi, for example, contain references to prices “dropping to mid-teens.” Wang introduced evidence that dozens of producers have participated in making the SIMM market, and that Mitsubishi sold $361 *1580million worth of SIMMs between April 1985 and March 1994. A reasonable person could conclude from the evidence that Mitsubishi supplied consideration to Wang by helping Wang achieve the market scenario it sought and valued.

Substantial evidence of the second and third benefits also appears on the record. Mitsubishi personnel testified that they made design changes to accommodate Wang’s needs: specifically, moving the capacitors onto the same side of the substrate as the memory chips. Similarly, testimony indicated that Mitsubishi declined to assess development and tooling charges to Wang because it understood Wang was permitting Mitsubishi to sell SIMMs to other customers. Clayton’s notes reveal that, from the first meeting onward, he closely followed Mitsubishi’s development and tooling of SIMMs with 64K or 256K chips in PLCCs. Clayton clearly wanted Mitsubishi to make 64K SIMMs, which Mitsubishi decided against, but Clayton also sought “to open [communication with] Japan for us on a 256K X9 SIMM.” All the while, Wang and Mitsubishi exchanged designs and samples. A reasonable person could conclude that Wang received something it valued because Mitsubishi was persuaded to follow Wang’s design suggestions, but was led to believe the resultant SIMMs were not custom products for Wang and did not assess charges. Although Wang points to contrary evidence, we cannot reevaluate credibility or substitute our choices for those of the jury. Perkin-Elmer, 732 F.2d at 893, 221 USPQ at 672-73. Therefore, these findings survive the substantial evidence test.

With the findings of fact firmly in hand, we now turn to the legal conclusion that an implied license exists. See Met-Coil Sys. Corp. v. Korners Unltd., Inc., 803 F.2d 684, 687, 231 USPQ 474, 476 (Fed.Cir.1986) (implied license is question of law). In patent law, an implied license merely signifies a patentee’s waiver of the statutory right to exclude others from making, using, or selling the patented invention. Spindelfabrik Suessem-Schurr Stahleclcer & Grill GmbH v. Schubert & Salzer Maschinenfabrik Aktiengesellschaft, 829 F.2d 1075, 1081, 4 USPQ2d 1044, 1048 (Fed.Cir.1987). In the words of the Supreme Court,

No formal granting of a license is necessary in order to give it effect. Any language used by the owner of the patent, or any conduct on his part exhibited to another from which that other may properly infer that the owner consents to his use of the patent in making or using it, or selling it, upon which the other acts, constitutes a license and a defense to an action for a tort.

De Forest Radio Tel. Co. v. United States, 273 U.S. 236, 241, 47 S.Ct. 366, 367, 71 L.Ed. 625 (1927). Since De Forest, this court and others have attempted to identify and isolate various avenues to an implied license. As a result, courts and commentators relate that implied licenses arise by acquiescence, by conduct, by equitable estoppel (estoppel in pais), or by legal estoppel. See AMP, Inc. v. United States, 182 Ct.Cl. 86, 389 F.2d 448, 452 nn. 4-5, 156 USPQ 647, 649 nn. 4-5 (1968); Robert L. Harmon, Patents and the Federal Circuit § 6.2(c) (3d ed. 1994 & Supp. 1996); 6 Ernest B. Lipscomb III, Walker on Patents § 20:14-20:17 (3d ed. 1987 & Supp. 1995). These labels describe not different kinds of licenses, but rather different categories of conduct which lead to the same conclusion: an implied license. The label denotes the rationale for reaching the legal result.

One of our predecessor courts observed that “courts generally have first looked for facts which give rise to an estoppel in the process of concluding that there is an implied license.” AMP, 389 F.2d at 452, 156 USPQ at 649. The opinions that hew most closely to the De Forest language and the “entire course of conduct” analysis rely on the doctrine of equitable estoppel, because De Forest requires that conduct of the patentee led the other to act. Bandag, Inc. v. Al Bolser’s Tire Stores, Inc., 750 F.2d 903, 925-26, 223 USPQ 982, 998-99 (Fed.Cir.1984); Stickle v. Heublein, Inc., 716 F.2d 1550, 1559, 219 USPQ 377, 383 (Fed.Cir.1983). In Bandag, we reversed a conclusion of implied license because the infringer failed to show an awareness of the conduct which supposedly created the license. Bandag, 750 F.2d at *1581925-26, 223 USPQ at 998-99. In Stickle, we affirmed that no implied license existed absent the required nexus between the patentee’s conduct and the infringing actions, given that the course of infringing action only began later and distinct from the cited conduct. Stickle, 716 F.2d at 1559, 219 USPQ at 383-84.

Neither this court nor the Supreme Court, however, has required a formal finding of equitable estoppel as a prerequisite to a legal conclusion of implied license. See AMP, 389 F.2d at 453-54, 156 USPQ at 651. To do so would remove all distinction between the doctrines. Rather the estoppel doctrines serve as guidelines. See id. at 451-52, 182 Ct.Cl. 86, 389 F.2d 448, 156 USPQ at 649-50. The primary difference between the estoppel analysis in implied license cases and the analysis in equitable estoppel cases is that implied license looks for an affirmative grant of consent or permission to make, use, or sell: i.e., a license. See Stickle, 716 F.2d at 1559, 219 USPQ at 383-84. Equitable estoppel, on the other hand, focuses on “misleading” conduct suggesting that the patentee will not enforce patent rights. A.C. Aukerman Co. v. R.L. Chaides Constr. Co., 960 F.2d 1020, 1041-44, 22 USPQ2d 1321, 1335-38 (Fed.Cir.1992). In Aukerman, we described a typical equitable estoppel situation as one in which (1) the infringer knows of the patent, (2) the patentee objects to the infringer’s activities, (3) but the patentee does not seek relief until much later, (4) thereby misleading the infringer to believe the patentee will not act. Id. at 1042-43, 960 F.2d 1020, 22 USPQ2d at 1335-36. Thus, the two doctrines- are not conterminous. Illustratively, both the jury, in its advisory capacity, and the district court decided against Mitsubishi on the defense of equitable estoppel per se, with misleading conduct as a required component of that defense.

Legal estoppel refers to a narrower cátegory of conduct encompassing scenarios where a patentee has licensed or assigned a right, received consideration, and then sought to derogate from the right granted. Spindelfabrik, 829 F.2d at 1080, 4 USPQ2d at 1048; AMP 389 F.2d at 452, 156 USPQ at 649-50. In AMP, for example, a patentee granted a license to, and received payment from, the United States for use of an “idea” which later became the subject matter of its patent. Id. at 453-54,182 Ct.Cl. 86, 389 F.2d 448, 156 USPQ at 651. The patentee then discovered a preexisting patent covering an aspect of the invention. Id. at 451, 182 Ct.Cl. 86, 389 F.2d 448, 156 USPQ at 648. After acquiring the preexisting patent, the patentee sued the government for infringement of the preexisting patent. Id. The Court of Claims held that an implied license barred patentee from using the preexisting-but-after-acquired patent to derogate from the express license negotiated under the other patent. Id. at 454, 182 Ct.Cl. 86, 389 F.2d 448, 156 USPQ at 651.

We review issues of law, like the implied license defense, de novo. Guinn v. Kopf, 96 F.3d 1419, 1421, 40 USPQ2d 1157, 1159 (Fed.Cir.1996). In this process, we review judgments rather than opinions. See, e.g., Baxter Healthcare Corp. v. Spectramed, Inc., 49 F.3d 1575, 1582, 34 USPQ2d 1120, 1125 (Fed.Cir.), cert. denied, — U.S. -, 116 S.Ct. 272, 133 L.Ed.2d 194 (1995). Moreover, we may affirm' a district court’s action where the record offers a route to affirm that neither expands nor contracts the rights established for either party by the judgment. Consolidated Aluminum Corp. v. Foseco Int’l Ltd., 910 F.2d 804, 815, 15 USPQ2d 1481, 1489 (Fed.Cir.1990). Here, we agree with the district court that the factual findings support the jury’s legal conclusion on the implied license defense, although we do not necessarily agree with all of the district court’s reasoning. The jury’s findings may support an implied license in the nature of legal estoppel given the transfer of a right for consideration and the subsequent suit for infringement, but we instead follow the lead of our predecessor, the United States Court of Claims, and focus on an alternative form of estoppel. See AMP, 389 F.2d at 452,156 USPQ at 650 (rejected equitable estoppel and affirmed based on legal estoppel).

Although judicially implied licenses are rare under any doctrine, Mitsubishi proved that the “entire course of conduct” between the parties over a six-year period led Mitsu*1582bishi to infer consent to manufacture and sell the patented products. See De Forest, 273 U.S. at 241, 47 S.Ct. at 367. Furthermore, the level of interaction between the parties while Mitsubishi designed and made SIMMs distinguishes this scenario from Bandag where the infringer based its failed defense on conduct unknown to the infringer when the infringement occurred. See Bandag, 750 F.2d at 925-26, 223 USPQ at 998-99. The record shows that Wang tried to coax Mitsubishi into the SIMM market, that Wang provided designs, suggestions, and samples to Mitsubishi, and that Wang eventually purchased SIMMs from Mitsubishi, before accusing Mitsubishi years later of infringement. We hold, as a matter of law, that Mitsubishi properly inferred consent to its use of the invention of Wang’s patents.

The findings that Wang bestowed “a right to use the SIMM invention” and that Mitsubishi supplied valuable consideration to Wang, support our holding that Wang’s conduct created a license. This falls short of the express licenses or assignments usually discussed in conjunction with legal estoppel, but it constitutes part of a course of conduct that transcends “unilateral expectations ... of one party.” Stickle, 716 F.2d at 1559, 219 USPQ at 383. Wang not only led Mitsubishi to infer consent, Wang obtained payment. Wang publicly announced its desired compensation. Wang also manifested this desire in other ways, for example, by Wang’s efforts at JEDEC to have its SIMM designated a standard. With the contributions of Mitsubishi and others, Wang received exactly the remuneration it desired: Wang’s design is an industry standard, and the benefits of a large market apd lower prices for SIMMs redound to this day. In sum, Wang consented to Mitsubishi’s use of the invention, granted the right to make, use, or sell the patented SIMMs without interference from Wang, and received consideration. Therefore, we agree with the district court’s determination, reiterated in denying Wang’s motion to amend the judgment, that Mitsubishi possesses an irrevocable royalty-free license under the ’513 patent.

Because the jury’s findings of fact are supported by substantial evidence and the jury reached the correct legal conclusion based on its findings, we affirm the judgment regarding Mitsubishi’s implied license defense. In reaching this result, we hold that Mitsubishi’s implied license is in the nature of equitable rather than legal- estoppel, because the license arose from an accord implicit in the entire course of conduct between the parties as discussed in De Forest and subsequent cases relying on equitable estoppel as a guideline. We further hold that the district court did not abuse its discretion by denying Wang a partial new trial on this issue.

Nonobviousness, Best Mode, & Infringement: ’513 Patent

The three issues raised in Mitsubishi’s cross-appeal remain for disposition. First, Mitsubishi asks us to reclassify several memory modules predating Wang’s invention as analogous prior art relevant to the nonobviousness of Wang’s SIMM, see Toshiba, 993 F.2d at. 863-64, 26 USPQ2d at 1773 (analogous means relevant to nonobviousness), and to hold the ’605 and ’513 patents invalid for lack of nonobvious subject matter. See Panduit Corp. v. Dennison Mfg. Co., 810 F.2d 1561, 1566, 1 USPQ2d 1593, 1595-96 (Fed. Cir.1987) (legal conclusion regarding nonobviousness relies on a factual foundation, including definition of the scope and content of prior art). In denying Mitsubishi JMOL on this issue, the district court summarized substantial evidence supporting the jury’s presumed- finding that the modules cited by Mitsubishi fall outside the scope of the analogous prior art. On appeal, Mitsubishi added nothing to detract from these findings. With the factual foundation intact, we affirm the conclusion on nonobviousness. See Panduit, 810 F.2d at 1566, 1 USPQ2d at 1596.

Next, Mitsubishi disputes the finding that Wang’s patents are not invalid for failure to disclose Clayton’s conceded best mode of practicing the invention. See 35 U.S.C. § 112 (1996). Noncomplianee with the best mode requirement involves a bifurcated factual inquiry. Great N. Corp. v. Henry Molded Prod., Inc., 94 F.3d 1569, 1571, 39 USPQ2d 1997, 1999 (Fed.Cir.1996). The factfinder must determine that subjec*1583of tively the inventor knew a better mode practicing the invention than the specification discloses, and that the inventor concealed the better mode. Id. The substantial evidence standard applies on appeal. Id. Although Wang admits that the specifications do not explicitly disclose Clayton’s contemplated best mode, Mitsubishi failed to convince the jury that Wang concealed Clayton’s best mode. The district court ruled that one skilled in the art would discern this best mode from the disclosure. After reviewing the record, we agree that substantial evidence supports the jury verdict.

Finally, Mitsubishi argues that the district court incorrectly interpreted Wang’s ’513 patent claims in upholding the infringement verdict. See Moleculon Research Corp. v. CBS, Inc., 793 F.2d 1261, 1269-70, 229 USPQ 805, 811 (Fed.Cir.1986) (claim interpretation is premise for factual comparison of claims to accused device). Courts must interpret disputed terms in patent claims, as a matter of law, based on the prosecution history, the specification, and the claims themselves. Markman v. Westview Instruments, Inc., 52 F.3d 967, 979, 34 USPQ2d 1321, 1329 (Fed.Cir.1995), affd, — U.S. -, 116 S.Ct. 1384, 134 L.Ed.2d 577 (1996). In its motion for JMOL, Mitsubishi interpreted five phrases in the ’513 patent’s claims. In seeking reversal, Mitsubishi reiterates its interpretation of those same claim elements.

Mitsubishi first asks us to interpret “a plastic leaded chip carrier” to refer solely to the “PLCC” made by Texas Instruments and used by Clayton in designing the SIMM. We hold that this language serves to distinguish the SIMM chip carriers from ceramic chip carriers found in the prior art, and that this language does not limit the claims to a particular make and model of plastic chip carrier. Second, Mitsubishi asserts that “substrate ... adequate for mounting thereon only in a single row said nine memory chips” limits substrate size to the bare minimum required for mounting nine chips in a single row. We hold that this language does not foreclose extension of the substrate, for some purpose, where the extra space cannot hold additional chips. Third, Mitsubishi contends that “thirty terminals for providing access” counts only terminals actively connected to a memory chip. We hold that “terminals” need not be connected. Fourth, Mitsubishi argues that “the ninth memory chip is for storing detection and correction information” means that the ninth chip must store two distinct kinds of information. We hold that this language does not necessarily refer to two types of information, especially given that Wang described the ninth chip as a “parity feature” storing only one kind of data known as “parity bits.” Fifth, Mitsubishi states that “support means for supporting the memory module at an angle” limits the claim to the style of pins used by Clayton in designing the SIMM. We hold that this language does not mandate a pin style matching Clayton’s initial style. Therefore, we arrive at the same interpretation as the district court. Since Mitsubishi does not deny that the claims as understood by the district court read on the accused modules, we affirm the finding of infringement.

AFFIRMED.

. Clayton mentioned that Wang’s legal office was investigating only the possibility of patenting the parity feature.

. Wang states that little evidence exists as to the exact number or content of meetings from December 1983 through 1989 but admits that the parties remained in contact and that from 1985 forward “about sixty meetings” took place, many addressing SIMM technology. Mitsubishi counters that Wang lost calendars and notebooks that would have filled in details of the discussions.

. As mentioned above,, the district court entered findings of fact on Mitsubishi’s equitable defenses, but not the implied license defense. We agree with Mitsubishi that some of the court's findings directly contradict the jury findings; where that occurs, we accept the jury's findings and consider the court's erroneous given that the court later upheld the jury findings in denying JMOL. See Newell, 864 F.2d at 762, 9 USPQ2d at 1421 (court’s fact finding subject to clearly erroneous standard); see also Therma-Tru Corp. v. Peachtree Doors, Inc., 44 F.3d 988, 995, 33 USPQ2d 1274, 1278 (Fed.Cir.1995) (juiy verdict operates as finding of fact precluding contrary judicial findings).