On June 8, 2008, the United States Supreme Court issued an important and far-reaching patent decision pertaining to the doctrine of patent exhaustion. [1] The decision strengthened the 150-year-old rule limiting a patent owner’s ability to partition product markets and to derive separate revenues for different uses of its patented technology by downstream purchasers.

The Supreme Court held in Quanta Computer, 128 S. Ct. 2109, 170 L. Ed. 2d 996, 2008 U.S. LEXIS 4702 (June 9, 2008), that the first sale or license of a patented invention by a component manufacturer also extinguished patent rights to the method of using a product that embodied or utilized the licensed component. The Court also clarified that the doctrine of patent exhaustion extinguishes patent rights to the invention when an incomplete component is licensed or sold, so long as that component “substantially embodies” the patented method or invention, and the only step necessary to practice the patent is application of common processes or the addition of standard components or parts. The Court refused to follow a formalistic approach to claim analysis and did not consider patent to be exhausted only when the device that practices the invention literally includes all patent claim elements. Finally, the Court also considered whether the doctrine of exhaustion was triggered by a license agreement that required a component manufacturer to warn purchasers that their use of the purchased component with other devices was not licensed or permitted by the patentee. The Court held that this arrangement could not avoid application of the doctrine of exhaustion because it failed to impose a “restricted use” contractual limitation on the component manufacturer, and therefore extinguished patentee’s right to assert patent infringement claims against downstream purchasers of the component.

While the decision has left some important questions unanswered, it is expected to have a significant impact on the structure of future license agreements and conduct of negotiations between many patent owners and component manufacturers of patented technology.

The Doctrine of Patent Exhaustion in the United States

The doctrine of “patent exhaustion,” also known as the “first sale” doctrine, was articulated by the U.S. Supreme Court over a century ago as a judicially created defense to charges of patent infringement. [2] This long-standing rule provides that the first authorized sale by the patent owner “exhausts” all patent rights to the product embodying the invention, thereby limiting a patentee’s rights in the invention and its ability to collect additional royalties after the first sale of the product. Under the doctrine, the patentee cannot demand, for example, additional royalty payments from a buyer who obtained the patented product from a licensed seller and subsequently decided to resell it.

The rationale for the rule is that once the first sale of the patented article has occurred, the patent owner has bargained for and obtained full payment for his patent rights that attach to the device embodying the patented invention. This rationale comports with the purpose and goal of the patenting system, the “main object” of which is the “promotion of the progress of science and the useful arts,” with the reward of inventors being a secondary concern and “merely a means to that end.” [3]

Over the last century, a growing number of exceptions to the “patent exhaustion” rule have come into existence. These exceptions stem from the principle that the doctrine of patent exhaustion does not apply to a sale or license that is expressly conditioned and limited to a particular use of the invention. In such a transaction, it is reasonable to infer that the parties negotiated a price that reflects only the value of the “use” rights conferred by the patentee, rather than the full reward for the rights to the invention. [4] For example, in the 1938 decision of General Talking Pictures Corp. v. Western Elec. Co., [5] the U.S. Supreme Court recognized the legality of restrictive patent licenses in general, and acknowledged that a patentee could avoid patent exhaustion by providing a narrow conditional license (also known as a “restrictive use” license) limiting the granted rights to a particular use of a licensed product. The patent license in General Talking Pictures authorized the licensee to make or sell amplifiers that embodied the patented invention strictly for home use purposes. [6] No license was given for any sales or use of the patented amplifiers in commercial movie theaters. The Supreme Court upheld patent infringement claims against unauthorized sales that went beyond the scope of the license grant, [7] and acknowledged that the unauthorized sale of the amplifiers to movie theaters did not trigger the doctrine of patent exhaustion. [8] By carefully restricting the patent license, the patentee in General Talking Pictures was thus able to partition the market and derive independent royalties from the sales of its patented invention to different market segments (home and commercial movie theater users).

Another example of a valid licensing restriction on a downstream use of a patented device was upheld by the United States Court of Appeals for the Federal Circuit in the 1992 decision of Mallinckrodt v. Medipart. [9] The label contract in Mallinckrodt involved a “single use only” licensing restriction on the patented device for delivery of radioactive or therapeutic materials. The patentee required each hospital-licensee to dispose of the patented device after each use. The licensee in Mallinckrodt did not follow this restriction and sent its used devices for reconditioning by a third party and subsequent reuse. The Federal Circuit acknowledged in Mallinckrodt that patentees are permitted to incorporate different types of conditions into their patent licenses to avoid patent exhaustion and to impose restrictions on the sale and use of their patented products by downstream purchasers. The court specifically upheld a “single use only” licensing restriction on the sale or use of a patented medical device, and confirmed a patentee’s right to maintain a patent infringement action against unauthorized resales. [10] The Federal Circuit contrasted “restricted use” licenses with “unconditional sales,” and appeared to conclude that only the latter type of arrangement exhausted patent rights. [11]

“Restricted use” licensing arrangements became increasingly common in many patent licenses, particularly after the Mallinckrodt decision. [12] By utilizing these restricted use licenses, patent owners and a large number of businesses were able to impose downstream restrictions on their inventions and to derive separate royalty revenues from downstream entities such as manufacturers, distributors and sellers of the devices that embody or employ the components produced by a licensed component manufacturer.

In Quanta Computer, [13] the U.S. Supreme Court addressed the scope and applicability of the “exhaustion doctrine” and the exceptions to the doctrine.

The Dispute in Quanta Computer v. LG Electronics

In Quanta Computer, the respondent, LG Electronics, purchased a number of computer technology patents, including the three patents at issue (collectively, the “LGE Patents”). The LGE Patents involved a system and method for optimizing performance and retrieval methodology for random access memory (RAM) used in a computer and a more efficient system and method for optimizing data transfer when multiple devices are connected to the same computer.

LGE licensed its patent portfolio, together with the LGE Patents, to Intel Corporation. The license arrangement permitted Intel to manufacture and sell microprocessors and chipsets that utilized the LGE Patents. In a separate agreement, however, Intel was required to give written warning to its own customers and purchasers concerning the scope of the LGE license. Intel had to warn purchasers of all its processors and chipsets that, while these Intel products were licensed by LGE and did not infringe its patents, the license did not extend, expressly or by implication, to any product that the purchasers made by combining an Intel product with any non-Intel product. [14]

The petitioners, including Quanta Computer, were a group of computer manufacturers (collectively, “Quanta”). They purchased microprocessors and chipsets from Intel, and then incorporated and installed them in their computer systems, in combination with a non-Intel computer memory and bus (a subsystem that transfers information among computer components). LGE filed a lawsuit against Quanta, alleging infringement of the LGE Patents by Quanta’s computer systems. The U.S. District Court for the Northern District of California initially ruled in favor of Quanta, holding that LGE’s patent rights were exhausted and that its potential patent infringement action was forfeited when it granted a patent license to Intel. In a subsequent order, the district court revised and limited its initial holding of exhaustion to patent claims in the LGE Patents that were directed to a physical device, such as an apparatus or system. The district court further modified its earlier decision by holding that the doctrine of patent exhaustion did not apply to the method or process patent claims.

On appeal, the Federal Circuit affirmed the district court’s determination that the doctrine of patent exhaustion did not apply to the method or process claims. In connection with the apparatus and system claims, the Federal Circuit reversed the district court and held that the doctrine of patent exhaustion was inapplicable because LGE never authorized or licensed Intel to sell its products for use in combination with non-Intel products. The Federal Circuit thus ruled that Intel could not have given Quanta an implied license under the LGE Patents because LGE had never given Intel a license in the first place to use or sell products that combined licensed Intel processors and chipsets with non-Intel components.

The Supreme Court Revives and Strengthens the Exhaustion Doctrine

The first question facing the Supreme Court in Quanta Computer was whether the first sale or license of the patented invention by LGE to a component manufacturer, namely Intel, also extinguished patent rights to the method of using a product that embodied or utilized the licensed component. The Supreme Court reversed the Federal Circuit’s decision that, because these claims could only be performed by the downstream purchaser and not by the component manufacturer, the exhaustion doctrine did not cover the method or process claims. The Supreme Court held that its earlier decisions dating back to the 1940s did not “differentiate transactions involving embodiments of patented methods or processes from those involving patented apparatuses or materials.” [15] The Supreme Court further reasoned that creating a broad rule that method claims as a category are nonexhaustible would seriously undermine the exhaustion doctrine and its practical applicability. If such rule were to be followed, a patentee would only need to include a method claim in each patent, together with or in place of an apparatus claim, to void the exhaustion doctrine altogether. [16] The Supreme Court acknowledged the danger of allowing such an “end-run” around exhaustion, and held that such a rule would “violate the longstanding principle that, when a patented item is once lawfully made and sold, there is no restriction on its use to be implied for the benefit of the patentee.” [17]

The second question addressed by the Supreme Court was to what extent a product must embody the patented invention in order to trigger patent exhaustion. Guided by a 1942 decision of United States v. Univis Lens, [18] the Supreme Court held that an incomplete computer component licensed to and sold by Intel “substantially embodies” the patented method “because the only step necessary to practice the patent is application of common processes or the addition of standard parts.” [19] This analytical approach shifted emphasis from the form and boundaries of the patent claims to the inventive part of each patented invention. More specifically, the Intel chips and products were designed to function only when the computer memory and bus were attached, and therefore were “lacking only the addition of standard parts” in order to practice the patented method. [20] The Court explained that the inventive part of the LGE Patents was not the fact that memory or buses were combined with an Intel microprocessor or chipset, even though such combination was literally required by the patent claims. Instead, everything inventive about each patent was embodied in the Intel products. [21] Furthermore, the Court found that the Intel chipsets had no reasonable noninfringing use. [22] Rather, the Intel components had only one reasonable and intended use, namely to practice the LGE Patents by connecting them to and working with the computer memory and bus. Therefore, the Supreme Court concluded that licensing and sale of the Intel products embodied LGE’s patented invention, including both the apparatus and method patent claims.

Having concluded that the method and apparatus claims in the LGE Patents were embodied in the Intel products, the Supreme Court then addressed the third issue: whether licensing restrictions imposed by LGE on Intel provided an exception and did not trigger the exhaustion doctrine. The licensing provision at issue was intended to skirt the exhaustion doctrine by expressly requiring Intel to warn purchasers of its components that their use of the purchased component with other devices was not licensed or permitted by the patentee. LGE had successfully argued before the Federal Circuit that Intel was never given an unrestricted license to sell patented products, and therefore could not convey implied license rights to others for the use of the Intel products in an unauthorized fashion with non-Intel products. The Supreme Court disagreed with this analysis, and drew a distinction between the implied license and exhaustion doctrine. The Court ruled that “exhaustion turns only on Intel’s own license to sell products practicing the LGE Patents.” [23] There were no restrictions in the Intel-LGE license agreement on Intel’s ability to make, use or sell the products manufactured by Intel and sold to purchasers who intended to combine them with non-Intel parts. A separate “master” agreement between Intel and LGE required Intel to give its customers written notice and warning regarding unlicensed combinations. The Court held that Intel was not restricted from making and selling components that substantially embodied the patented invention, and that Intel’s license and rights to sell were not conditioned on the notice to Quanta or on Quanta’s decision to abide by LGE’s warning concerning unauthorized use. Because of the lack of those contractual restrictions in the LGE-Intel license, the doctrine of patent exhaustion prevented LGE from further asserting its patent rights against Quanta and downstream purchasers of Intel products.

The Impact of the Quanta Computer Decision

The Supreme Court’s ruling in Quanta Computer resolved a couple of important questions concerning the exhaustion doctrine. First, it foreclosed two easy routes that patentees might have used to escape patent exhaustion. A patentee can no longer shield its patent from the doctrine of exhaustion by simply adding method claims, or by securing another related patent with method claims and directing those method claims to cover products used by the downstream purchasers.

Second, the Supreme Court clarified that patent exhaustion is triggered when there is first sale of a product that “substantially embodies” the inventive features of the patent, even when the patent claims include some additional common elements or steps. Patentees can no longer avoid the doctrine of exhaustion by on the one hand selling and licensing products that embody the core inventive features, while on the other hand requiring in the patent claims some additional well-known steps or common components that need to be added in order to make the invention complete or to practice the patented method.

In the litigation context, however, the Quanta Computer decision has still left some room to argue against patent exhaustion. For example, in connection with so-called “combination patents” (claiming the novel combination of existing components), it might still be argued that a component product does not “substantially embody” the invention because all elements or components need to be combined together in order to constitute an inventive feature. It might also be argued for some patents that exhaustion does not apply because the licensed component has some “reasonable use” that does not practice the patent. In sum, the legal test set forth by the Supreme Court in Quanta Computer leaves to a case-by-case determination the qualification and quantification of what constitutes “substantial” embodiment of each patented invention.

Third, the Supreme Court was expected to resolve in Quanta Computer whether the Federal Circuit’s decision in Mallinckrodt was still good law, and whether the “restricted use” licenses could ever be utilized to impose restrictions on the downstream sales of patented inventions. Instead, the Supreme Court sidestepped the broader issue, and narrowly interpreted the license between LGE and Intel to resolve the issue in Quanta Computer, without drawing any broader generalities or conclusions regarding use of “restricted use” licenses to partition a product market. The Court interpreted the license between LGE and Intel as an “unrestricted” type of license, which triggered the patent exhaustion rule. The Court declined to extend this holding to or offer an opinion regarding circumstances involving “restricted use” licenses.

Fourth, the Quanta Computer decision not only signifies important changes in the application of the doctrine of exhaustion, but also has a significant impact on the way in which patent owners and licensors should structure their license agreements and conduct licensing negotiations. If a licensor is interested in securing a “restricted use” license, and would like to be able to assert patent infringement claims against downstream purchasers, it must now clearly impose the restrictions on both the licensee and, through that licensee, on downstream purchasers. In other words, the patent owner must now retain the right to sue directly for patent infringement the licensees and component manufacturers that sell patented products or components for a field of use that goes beyond the narrow restrictions in the license agreement. This potential exposure to a direct infringement lawsuit is not something that many component manufacturers would willingly accept. Furthermore, if such conditions were to be imposed, many component manufacturers might insist on reducing existing royalties that they would be willing to pay for a narrow and circumscribed patent license that allowed a direct patent lawsuit against them. Imposing such express contractual restrictions on licensees and component manufacturers might also carry an added burden of monitoring the intended use of the licensee’s or component manufacturer’s products by downstream purchasers. For example, if the license granted to Intel in Quanta Computer did not permit manufacture and sale of the Intel processors and chipsets for use with a particular type of computer architecture, then Intel might choose to monitor the intended use of its products by downstream purchasers in order to limit its exposure to a direct patent infringement action by LGE. This type of monitoring is both expensive and time-consuming, and many component manufacturers are likely to shift this added burden to the patentee, and thus reduce royalty payments accordingly.

Finally, because the Supreme Court did not fully resolve in Quanta Computer the issue of legality and enforcement of the “restricted use” licenses in general, there remains some uncertainty as to whether a patent owner could enforce its patent rights even if it were successful in negotiating a “restricted use” license with a component manufacturer.