Quanta Computer, Inc. v. LG Electronics, Inc. (2008)

    By Kevin E. Noonan --

The Supreme Court handed down its opinion in Quanta Computer, Inc. v. LG Electronics, Inc. today, deciding a question of patent exhaustion in the sale of computer components.  The decision, strengthening the patent exhaustion principle, was not really a surprise for a number of reasons.  First, it was consistent with Supreme Court precedent, most particularly United States v. Univis Lens Co., 316 U. S. 241 (1942).  Second, the decision is in keeping with the Supreme Court's longstanding antipathy to patent "monopolies" and anything that would result in extending them outside their most limited scope.  Finally, of course, the case represented another opportunity to reverse the Court of Appeals for the Federal Circuit; the Supreme Court has spent the last decade or so in establishing its own supremacy in patent law at the expense of whatever consistency the Federal Circuit has attempted (rightly or wrongly) to impose.

The case involved sales by Quanta and the other defendant computer makers of computers comprising both Intel and non-Intel produced components.  LG Electronics, who licensed the patents-in-suit (U.S. Patent Nos. 4,939,641; 5,379,379; and 5,077,733) to Intel, imposed a condition on its licenses with Intel that sales were not authorized to third parties that would mix Intel and non-Intel components, and further that Intel was under an affirmative obligation to inform its customers that such uses were not licensed.  The patents-in-suit claimed methods for organizing read/write requests in computer cache and random access memory (the '379 patent); methods for prioritizing access of peripheral devices to the microprocessor via the computer bus (the '733 patent); and a system for "ensuring that the most current data are retrieved from main [RAM] memory by monitoring data requests and updating main memory from the cache when stale data are requested" (the '641 patent).  It was undisputed that Quanta and the other defendants purchased patented microprocessors from Intel that fell within the scope of the LG patents, that they were aware of the restrictions and limitations placed on their use of these components by the license between LG and Intel and that the defendants sold computer systems comprising Intel components operatively linked to non-Intel components.  Although the District Court granted summary judgment in favor of Quanta, the Federal Circuit reversed in part and remanded, on the grounds that method claims were not subject to patent exhaustion or, in the alternative, that LG had not licensed Intel to sell components for use in computers having both Intel and non-Intel components.

In its decision reversing the Federal Circuit, a unanimous Court (in an opinion written by Justice Thomas) settled one unsettled question of patent exhaustion and reinforced its test for calculating the extent of the doctrine.  Directly reversing the Federal Circuit, the Court held that method claims are not exempt from patent exhaustion.  Citing previous Supreme Court cases that "rest on solid footing" (including Ethyl Gasoline Corp. v. United States, 309 U.S. 436 (1940), and Univis), the opinion asserted that it had never treated composition and method claims differently, applying the "substantial embodiment" test to apply the patent exhaustion principle to method claims.  The reasoning mirrored in some ways the District Court's reasoning in University of Rochester v. G. D. Searle on the composition/method distinction:  if method claims were not subject to exhaustion, all patents would contain (or all patent families would contain) method claims that could be used to coerce royalty payments otherwise exhausted for composition claims by sales of a patent-protected item.  The Supreme Court (reasonably) concluded that patent exhaustion must apply to method claims to prevent such shenanigans (characterized by the Court as an "end-run around exhaustion") from negating the value of the doctrine to the public. 

The Supreme Court reiterated the test for whether exhaustion applies that it enunciated in Univis:  sales that "substantially embody" the patents in suit will be sales that exhaust the patent right to obtain royalties.  The Univis case involved sales of eyeglass "blanks" for making patented compound lenses, and the products sold by the alleged infringers were the final eyeglasses containing lenses ground from the blanks sold by the patentee.  In Univis, the Court opined (albeit in an action brought by the United States under the antitrust laws, not the Patent Act) that the lens blanks embodied essential features of the patented lenses and thus sales of the blanks exhausted patent rights in the finished lenses. 

Here, the Supreme Court concluded that the components at issue, microprocessors and chipsets, "substantially embodied" the system and method claims of the patents-in-suit.  It analogized these components with the lens blanks in Univis, which it asserted "did not fully practice the patents at issue" because they needed to be ground into lenses, here the patented microprocessors and chipsets did not "function at all" until they were assembled into the final product (i.e., the computer comprising non-Intel components).  "If [the opinion states], as in Univis, patent rights are exhausted by the sale of the incomplete item, then LGE has no postsale right to require that the patents be practiced using only Intel parts."  The Court rejected the three LG arguments that attempted to distinguish this case from Univis.  First, according to LG, Univis is properly limited to cases where the products "contain all the physical aspects needed to practice the patent."  Second, this case can be distinguished from Univis because there was no "patentable distinction" between the lenses and lens blanks in Univis, and there are such distinctions here (inter alia, the existence of multiple patents for the alleged infringing articles).  Finally, LG invoked Aro Mfg. Co. v. Convertible Top Replacement Co., 365 U.S. 336 (1961) for the countervailing principle that "combination" patents should not be exhausted by the sale of one component of the combination. 

On the contrary, the Supreme Court found Univis controlling.  The "only reasonable and intended use [of the microprocessors and chipsets] was to practice the patent," said the Court, just as the only "reasonable and intended use" of the lens blanks in Univis was to make the patented eyeglasses.  Moreover, the microprocessors and chipsets embodied the essential features of the patented invention, since all that was required was to attach them to other conventional computer components in conventional ways (much as the lens blanks in Univis were converted to lenses using conventional grinding techniques not encompassed by the patent claims).  (Interestingly, the Court noted that Quanta was limited to using such conventional methods because the internal structure of the Intel microprocessors and chipsets were protected by Intel as trade secrets, and Quanta could not modify them.  "Intel all but practiced the patent itself by designing its products to practice the patents, lacking only the addition of standard parts.")  Finally, the Court rejected LG's reliance on Aro, distinguishing that case as involving replacement of a part of a combination and because Aro is limited to claims relying on the combination of elements and not any individual component of the combination.

Finally, the Supreme Court addressed the question of whether the license terms precluded application of the exhaustion doctrine and determined they did not.  Important in the Court's determination was that the license agreement did not contain the preclusive condition; a separate "Master Agreement" between the parties recited that provision.  But "Intel's authority to sell its products embodying the LGE Patents was not conditioned on the notice [that the mixed combination of components was an unauthorized use] or on Quanta's decision to abide by LGE's directions in that notice," because breach of the provisions of the Master Agreement precluding sale of Intel microprocessors and chipsets to third parties who combined them with non-Intel components did not constitute a breach of the licensing agreement.

The decision has important ramifications for biotechnology patent claims.  There are two immediately-evident examples:  cases where a patentee having method claims and composition claims limits by so-called "label license" the uses for a patented article; and cases where a patented article has the biological property of replication, where the license precludes use of replicates of the article after purchase.  An example of the first type of situation are the limits placed on the practice of the polymerase chain reaction, based on patents to the amplification method and the thermostable polymerase.  The restrictions were of two types:  first, licenses to the method were granted only upon purchase of the polymerase and use of an "authorized" thermocyler.  Second, the method (and for that matter, the polymerase) was not licensed for diagnostic uses, only for scientific research.  Under the patent exhaustion doctrine set forth in the Quanta opinion, it is likely that neither of these restrictions would be enforceable.  First, although the polymerase has other uses that would not infringe the claims of the PCR method patent, it is likely that sale of the thermostable polymerase would exhaust the method claims as well, since the thermostable characteristic of the polymerase embodies essential features of the claimed invention.  This is even more likely regarding the diagnostic use prohibition, since it represents the kind of restriction the Court prohibited in Adams v. Burke:  "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.'"  Adams, 17 Wall. 453, 457 (1873).

The second type of post-sale restrictions important to biotechnology are those that restrict use of a patented article that is capable of self-replication; the most (in)famous cases of this type are the Monsanto herbicide-resistant seed cases, where the "label license" prohibits replanting seed produced using the recombinant seed purchased from the company.  Monsanto has recently been the victor in a number of lawsuits against farmers who were found to have harvested recombinant seed from one years' planting and replanted it the following year; the Federal Circuit has upheld lower court determinations that this constituted patent infringement as activity falling outside the scope of the license (see, e.g., Patent Docs post on Monsanto Co. v. David).  Although critically important economically (since it is clear that Monsanto would not be in the seed business for very long in the absence of these restrictions), it is difficult to see how these restrictions could pass muster under the rubrics for applying patent exhaustion set forth in today's Quanta decision.  Certainly, it is without question that the seeds "embody" (figuratively and literally) the "essential features" of the patented invention, and thus the types of limitations Monsanto (and other seed producers) have placed on replanting of their patented seeds are unlikely to prevail if they try to assert them under the patent laws.

As has been the Supreme Court's penchant in several recent cases, there is an important caveat to its holding to be found in a footnote; here, it is footnote number 7:

We note that the authorized nature of the sale to Quanta does not necessarily limit LGE's other contract rights.  LGE's complaint does not include a breach-of-contract claim, and we express no opinion on whether contract damages might be available even though exhaustion operates to eliminate patent damages.  See Keeler v. Standard Folding Bed Co., 157 U. S. 659, 666 (1895) ("Whether a patentee may protect himself and his assignees by special contracts brought home to the purchasers is not a question before us, and upon which we express no opinion.  It is, however, obvious that such a question would arise as a question of contract, and not as one under the inherent meaning and effect of the patent laws.").

Monsanto also employs a contract to prohibit replanting of its seeds, and it is clear that a patentee could condition sale of a patented article on contractual provisions restricting its use.  Such an arrangement has clear disadvantages as compared with patent law claims, including variations in application of the law (and enforceability of the agreement) state-by-state, as well as the potential to implicate the antitrust laws.  But such agreements may be the only way remaining for biotechnology companies to protect their inventions in both the preceding types of situations.

The Quanta decision provides comfort to those many who have been battling seed producers over replanting restrictions.  It is also yet another example of the value of patent rights being eroded by Supreme Court decisions that run counter to the determinations of the Federal Circuit.

Comments

[Jeff]

Isn't it distinguishable that the chip sets in LG and the seeds are quite different? For argument sake, the farmer can do whatever he/she wants with the original seeds he bought from the company. But the moment those seeds are planted, patentee's rights are exhausted and so does the farmer's. The resulting seeds are "new" and amounts to unlicensed copying. If Intel paid for one chip but uses the licensed chip's components (albeit by destroying) to make a second chip, would that be permissible under the exhaustion doctrine? Just a thought....

[kk]

Mr. Noonan,

I am a little confused by your characterization of the LGE-Intel license in the second paragraph.

My understanding of the SCOTUS opinion is that LGE did not impose any condition on Intel to limit sales. Intel was required under a separate agreement to notify its customers that mixing Intel and non-Intel chipsets would not be covered under the LGE-Intel license. However, LGE did not require Intel to limit sales to only those customers who agreed not to mix Intel and non-Intel components.

Am I incorrect in my reading? Thanks.

[Kevin E. Noonan]

Dear KK:

You are correct. Perhaps in an effort to be succinct I was unclear, but I think this passage embodies your understanding:

"Important in the Court's determination was that the license agreement did not contain the preclusive condition; a separate "Master Agreement" between the parties recited that provision. But "Intel's authority to sell its products embodying the LGE Patents was not conditioned on the notice [that the mixed combination of components was an unauthorized use] or on Quanta's decision to abide by LGE's directions in that notice," because breach of the provisions of the Master Agreement precluding sale of Intel microprocessors and chipsets to third parties who combined them with non-Intel components did not constitute a breach of the licensing agreement."

So, Intel's license from LG did not preclude LG from selling the patented chipsets to a third party regardless of whether the 3rd party complied with the provisions of the Master Agreement, and breach of those provisions was not considered breach of the patent license. This is "eating your cake and having it too" behavior; I'm not sure the outcome is different if these facts were different, but they provided further justification for the Court to rule the way it did.

Thanks for the comment. Hope this clarifies what I intended to say.

[Kevin E. Noonan]

Dear Jeff:

I think the distinction you draw is certainly Monsanto's argument. The portion of the opinion that cuts against it is where the Court talks about how attaching the patented chipsets to the bus and peripherals was the intended use of the patented products, which had "no function" until you made the attachments. Sounds like they would say that planting the seed is the "intended use" and producing seeds is what plants do. It isn't like someone took the chipsets and copied them - it's more like the chipsets automatically producing a backup copy of the information (or better, logical structure) of the chip and then transferring this onto a naive chip. But even that would require an additional step (providing the naive chip) that doesn't apply to seeds.

And good luck getting the CAFC to reverse a District Court that applies Quanta to a seed case the way I postulate.

Thanks for the comment.

[Anon E. Mouse]

Jeff & Kevin,

That's how I read the Monsanto cases--the infringing (second and later generation) seeds were never sold. Only the seeds in the dirt were sold, and exhaustion attaches only to the sold item itself.

Even under Quanta, there is no authorized sale of the second and later generation seeds, and exhaustion doesn't attach to the object.

A better Quanta analogy would be if the accused infringer used the purchased chipset to make a computer-driven foundry that cranks out infringing chips.

The making of the computer-driven foundry would not be an infringement, but cranking out infringing chips would be.

My guess is that the Fed Cir would reverse a district court over this...unless it can find something in the license agreement to wiggle out of confronting the issue.

(and I don't mean "wiggle" in a bad way, that's what they should do).

The Monsanto cases will be a very big deal now, especially with the seed folks openly promising to increase yields for "green" energy + food. They won't do it without patent protection, and contract won't hack it (privity).

[Anon E. Mouse]

clarification on my analogy:

my chipsets' only substantial use was in making a chipset foundry. No naive-chip step required.

[Chris Newman]

Here's where I see some potential ambiguity in the decision, and would like to know what others think.

It seems that there are two requirements for exhaustion:

1) The sold article is capable of use only in practicing the patent.
2) The sold article embodies essential features of the patented invention.

Now, in applying 2) on the facts of this case, the Court highlights the fact that ALL the inventive aspects of the patent were already contained in the chipsets. (Query: is the Court essentially saying that the PTO was wrong to issue the method claims as a separate patent? Didn't LGE argue in its briefs that the PTO must have determined the method to contain inventive aspects beyond what was already contained in the chipsets?) So there would appear to be room for future owners of method patents to argue that those patents are not exhausted by sale of an article of which 1) is true, and that embody SOME essential features of the invention, so long as there are other essential features of the invention that are not embodied in the article.

Say for example that LGE’s chipsets came in two flavors, A and B. Each embodies a different essential feature of the patent, and to work, you have to combine an A with a B (and a bus and memory). So what if LGE makes unrestricted sales of A sets, but keeps the authorization to sell B sets subject to strict license restrictions? Can it still enforce its method patent against purchasers of A who practice the method in violation of the restrictions on B sets?

Or what if the B component isn’t independently patented at all, but the idea of combining it with A to achieve the desired result was inventive? Can the method patent still be enforced against purchasers of A?

[Kevin E. Noonan]

Dear Mouse:

If the seeds were not "sold," then infringement would lie for making and using, and that sounds reasonable.

However, you are then stuck with characterizing two types of seeds, which are identical except for when (and by whom) they were produced: seeds purchased and seeds made by plants from the purchased seeds. Although I see the distinction, the problem is that the Court has made a bright line rule for cases that are distinguishable. And since growing and producing more seed is an essential feature of the plants produced with the recombinant seed (and remember, the other aspect is that those plants are herbicide resistant), the plants have to be grown the same way the chipsets had to be attached to other components.

The infringing act would then have to be the second planting with the unsold seeds. But I'm not sure this distinction will have much weight with the Court, particularly in view of its statements that the patent laws do not exist to make inventors wealthy, and because of the herbicide sales piece. Since the second planting will only confer an advantage if the farmer buys the herbicide from the company, this may be a situation where the equities don't lie with the patentee. Recall the equitable flavor of the part of the opinion regarding (non)breach of the license by Quanta's using non-Intel components. I just see the Court going the same way on this question.

Thanks for the comment.

[Kevin E. Noonan]

Dear Chris:

Your hypotheticals are interesting, but the biggest problem is that the Court made it clear that clever isn't good enough. The Court has had a disinclination to protect patent rights that stems from a 19th Century mentality regarding technology, a turn of the (last) century antipathy towards "monopolies," and a middle of the last century notion that patents are anticompetitive. Since the Court sees itself in historical continuity, its decisions are often bogged down with outdated notions (think, "separate but equal" for example).

It will take something more than clever to nudge the Court from this mindset, and figuring out ways to increase licensing fees for patentees, in my opinion, won't be enough. Now, prove that seed companies won't have the capital to do the research that will increase harvests enough to prevent Americans from starving, well, THAT may be enough. But who wants that?

Thanks for the comment.

[e6k]

(Query: is the Court essentially saying that the PTO was wrong to issue the method claims as a separate patent? Didn't LGE argue in its briefs that the PTO must have determined the method to contain inventive aspects beyond what was already contained in the chipsets?)

Yeah I picked up on that too. But the cout is right, and this is one reason why I believe the decision to let people claim the same invention as a method and a product is horrendously bad. Further, the problem with that situation runs deeper into the ability to distinguish method claims by the structural features of the articles on which they are performed or create. Product claims for products, method claims for methods I say. You'll notice that this approach substantially invalidates a huge portion of issued patents. Even so, it's the right thing to do.

"Say for example that LGE’s chipsets came in two flavors, A and B. Each embodies a different essential feature of the patent, and to work, you have to combine an A with a B (and a bus and memory). "

Now you're making allegations that neither A nor B are enabled unless both are disclosed together. I agree with this as well, it is through carefull drafting etc. that lawyers split this crp up and burden the office so that their clients can enjoy broader protection.

[e6k]

ahem, excuse me not "enabled" but "not the best mode" instead. Though, come to think of it, enabled might be sufficient to describe that situation since if you don't know how to make B then you don't really know how to practice A.

[Anon E. Mouse]

Dear Kevin,

You may be correct, but I think we're at odds on the equities.

As I read Quanta, Intel originally bargained and paid for the right to make, use, and sell the chips without restriction. It was in a separate (later, IIRC) Master Agreement that required Intel to give notice to the buyers.

In Monsanto, both sides bargained for the no-replant clause which was incorporated into the price of the seeds.

As I see it, the equities are with Monsanto.

[Kevin E. Noonan]

Dear e:

But of course you recognize that the reason there are separate patents for methods and compositions is usually an Office-imposed restriction requirement (at least in the chemical/biotech arts). So although I agree with you that in the hypo where A+B are needed to be functional a claim to A or B alone may be inoperative, the fact that a patentee has multiple patents to the same "invention" can't be a basis for invalidating them.

Thanks for the comment.

[Kevin E. Noonan]

Dear anon:

You may be right about the equities, and that may work in a breach-of-contract action (unless the court were persuaded that it was really an adhesion contract, since Monsanto's patent position gives it a monopoly on the recombinant seeds).

But all the Supreme Court seemed to be doing here was 1) reestablishing the principle of patent exhaustion if it had been diminished by the CAFC; 2) excluding patent infringement as a cause of action in this kind of dispute and 3) (of course) bringing the CAFC into line by exerting it supremacy. I think the case might have been different if Intel's license had been predicated on preventing non-Intel components from being used with Intel's microprocessors and chipsets. If it was a field of use license, for example, or if Intel warranted not only that it would inform but also police compliance (so that its sales were conditional and if the condition were not met Intel would be the plaintiff), it might have been a different case.

Thanks for the comment.

[Alex Jones]

I rarely do not comment on blogs but yours I had to stop and say Great article!!

[Zubair]

Hi,

That is very informative and well written. Mentioned below is an excerpt from an article on Patent exhaustion in India:

"A patent grants the Patent holder exclusive rights to prevent others from making, using, selling, offering for sale in the territory of patent grant or importing an invention into the territory of patent grant. Once an unrestricted sale of the patented invention is made, the rights of the patent holder with respect to the product are exhausted and this is called as the Doctrine of Exhaustion or First Sale Doctrine......to read more please visit http://indianipinfo.blogspot.com/2010/10/patent-exhaustion-in-india.html