ORDER RE CLAIM CONSTRUCTION AND INVALIDITY Re: Dkt. No. 158
JAMES DONATO, District Judge.
In this patent infringement case, Altera seeks a declaration from the Court that it does not infringe certain patent claims asserted against it by PACT, and that those claims are invalid. As is usual in patent cases, the Court is charged with the task of determining the meaning of various claim terms over which the parties disagree. See Markman v. Westview Instruments, Inc., 52 F.3d 967, 978-79 (Fed. Cir. 1995) (en banc), aff'd, 517 U.S. 370 (1996). The Court has also allowed the parties to brief whether two claim terms are indefinite, rendering the claims that contain them invalid. This order addresses claim construction and Altera's motion for summary judgment on indefiniteness.
The order follows full briefing on these issues and a hearing on claim construction. In addition, the Court has appointed Professor John Wawrzynek as a neutral expert under Federal Rule of Evidence 706, and invited him to submit any comments he had on claim construction, which he did. See Comments of John Wawrzynek ("Wawrzynek Comments"), Dkt. No. 220-1. Altera moved for leave to file a response to those comments, attaching its proposed response, see Dkt. No. 188, and the Court granted the motion while allowing PACT to file a response. See Dkt. Nos. 192, 197. The Court has considered both Professor Wawrzynek's comments and the parties' responses, as well as the declarations submitted by the experts retained by the parties.
PACT asserts six patents, which fall into three patent families:
See Preliminary Election of Asserted Claims, Dkt. No. 54.
Generally speaking, the patents-in-suit relate to reconfigurable semiconductor devices, such as field programmable gate arrays ("FPGAs"). Unlike application specific integrated circuits ("ASICs"), which are manufactured to perform specific design tasks, reconfigurable devices can be reprogrammed to suit various applications after manufacturing. Most reconfigurable semiconductor devices have a multi-dimensional programmable cell architecture, meaning they are composed of a matrix of configurable logic blocks known as cells or processing array elements ("PAEs"), which are connected to one another.
The bus system patents are directed at providing a "uniform" or "general" bus system. See '077 patent, 1:57-59, 1:64-66. According to the patents, prior art systems required a programmer who wanted to interconnect processors or connect peripherals to the processors to integrate functionality specific to those devices into the cells. See id. at 1:38-44. That not only increased wiring complexity, but also took up additional space on the unit. See id. at 1:46-67. The "uniform" bus system described by the bus system patents attempts to create a general-purpose bus system that allows "[a]ny number of memory devices, peripherals or other units" to be connected. See id. 1:64-2:3.
The dynamic reconfiguration patents involve devices that consist of an array of cells. See '869 patent, 1:63-2:1. The patents describe allowing certain cells to be reconfigured without affecting the neighboring cells, see id. 2:19-26, as distinguished from prior art systems, which allegedly required reconfiguring all of the cells at one time, thus interrupting even those cells that were not being reconfigured. See id. at 1:28-43.
Finally, the '525 patent is directed at an improved cell. See '525 patent, 2:37-39. The allegedly inventive cell includes an "expanded arithmetic and logic unit" ("EALU"). See id. 2:36-41. According to the specification of the '525 patent, the allegedly inventive EALU reduces the volume of data necessary for configuration, see id. at 2:41-43. It thus improves on prior art units, which took much more space, time, processing power, and money to reconfigure cells. See id. at 1:31-43.
A. Claim Construction
Claim construction is a "mongrel practice." Markman, 517 U.S. at 378. Although the "ultimate issue of the proper construction of a claim should be treated as a question of law," it can involve subsidiary factual findings. See Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 135 S.Ct. 831, 838 (2015).
By default, claim terms are given their "ordinary and customary meaning," which is "the meaning that the term would have to a person of ordinary skill in the art in question at the time of the invention." Phillips v. AWH Corp., 415 F.3d 1303, 1312 (Fed. Cir. 2005) (en banc). There are only two circumstances where a claim is not entitled to its plain and ordinary meaning: when a patentee sets out a definition and acts as his or her own lexicographer, or when the patentee disavows the full scope of the claim term either in the specification or during prosecution. Thorner v. Sony Computer Entm't Am. LLC, 669 F.3d 1362, 1365 (Fed. Cir. 2012).
When construing claim terms, the Federal Circuit emphasizes the importance of evidence that is intrinsic to the patent: the language of the claims themselves, the specification, and the prosecution history. Phillips, 415 F.3d at 1312-17. The claim language can "provide substantial guidance as to the meaning of particular claim terms," both through the context in which the claim terms are used and by considering other claims in the same patent. Id. at 1314. The specification is likewise a crucial source of information: Although it is improper to read limitations from the specification into the claims, the specification is "the single best guide to the meaning of a disputed term." Id. at 1315 ("[T]he specification `is always highly relevant to the claim construction analysis. Usually, it is dispositive. . . .'"); see also Merck & Co. v. Teva Pharms. USA, Inc., 347 F.3d 1367, 1370 (Fed. Cir. 2003) ("[C]laims must be construed so as to be consistent with the specification. . . .").
Despite the importance of the intrinsic evidence, courts may also use extrinsic evidence — technical dictionaries, learned treatises, expert and inventor testimony, and the like — to help construe the claims. Phillips, 415 F.3d at 1317-18. When considered in conjunction with the intrinsic evidence, dictionaries may reveal what the ordinary and customary meaning of a term would have been to a person of ordinary skill in the art at the time of the invention. Frans Nooren Afdichtingssystemen B.V. v. Stopaq Amcorr Inc., 744 F.3d 715, 722 (Fed. Cir. 2014) (citing Phillips, 415 F.3d at 1312-18) ("Terms generally carry their ordinary and customary meaning in the relevant field at the relevant time, as shown by reliable sources such as dictionaries, but they always must be understood in the context of the whole document—in particular, the specification (along with the prosecution history, if pertinent)."). Extrinsic evidence is, however, "less significant than the intrinsic record in determining the legally operative meaning of claim language." Phillips, 415 F.3d at 1317 (internal quotations omitted).
Patent claims must "particularly poin[t] out and distinctly clai[m] the subject matter which the applicant regards as his invention." 35 U.S.C. § 112 ¶ 2. A claim fails to satisfy this requirement and is invalid if its language, when read in light of the specification and the prosecution history, "fail[s] to inform, with reasonable certainty, those skilled in the art about the scope of the invention." Nautilus, Inc. v. Biosig Instruments, Inc., 134 S.Ct. 2120, 2124 (2014).
Patents are entitled to a presumption of validity. As a result, any fact critical to a holding on indefiniteness must be proven by the challenger by clear and convincing evidence. See Intel Corp. v. VIA Techs., Inc., 319 F.3d 1357, 1366 (Fed. Cir. 2003); Nautilus, 134 S. Ct. at 2130 n.10 (declining to address standard of proof for facts underlying conclusion of indefiniteness). However, "this presumption of validity does not alter the degree of clarity that § 112, ¶ 2 demands from patent applicants." Nautilus, 134 S. Ct. at 2130 n.10.
I. CLAIM CONSTRUCTION
C. "at least one of . . . and . . ."
The first of the parties' disputes is one of syntax, not semantics. Several claims use a grammatical construction of the form, "at least one of A, B, and C." Altera says such constructions should be interpreted to mean "at least one of A, and at least one of B, and at least one of C." By contrast, PACT argues that they should be read to mean "at least one of A, and/or at least one of B, and/or at least one of C," or equivalently, "at least one element chosen from the set consisting of As, Bs, and Cs."
It is inarguable that Altera's reading of this construction was endorsed by the Federal Circuit in SuperGuide Corp. v. DirecTV Enters., Inc., 358 F.3d 870, 886 (Fed. Cir. 2004), which created a presumption that "at least one of" applies to each element of the following list, and that the word "and" be read conjunctively. And properly so, because as PACT concedes, it could have unambiguously conveyed the meaning it now urges by using "and/or" in place of "and." But it is also inarguable that district courts have often bucked SuperGuide's presumption where called for by context or other intrinsic evidence. See Fujifilm Corp. v. Motorola Mobility LLC, No. 12-cv-03587-WHO, 2015 WL 1265009, at *7-9 (N.D. Cal. Mar. 19, 2015) (collecting cases); Radware Ltd. v. A10 Networks, Inc., Nos. 13-cv-02021-RMW, 13-cv-02024-RMW, 2014 WL 1572644, at 6-7 (N.D. Cal. Apr. 18, 2014) ("SuperGuide has not been read as a uniform rule that `at least one of . . . and' be construed in the conjunctive."); Power-One, Inc. v. Artesyn Technologies, Inc., No. 05-cv-00463, 2007 WL 896093, at *14 (E.D. Tex. Mar. 22, 2007) (distinguishing SuperGuide and construing "at least one of X, Y, and Z" as "a group [of X, Y and Z from which] at least one is selected").
This flexibility is well warranted. SuperGuide itself suggested that intrinsic evidence could override its default rule, pointing out that every disclosed embodiment of the patent was consistent with its reading of the claims, while at least one would be inoperable under the alternate proposed reading. SuperGuide, 358 F.3d at 887-88. That is because the twin canons on which SuperGuide relies — that "at least one of" applies to each element of the following list and that "and" must be read conjunctively — are meant to be practical guidelines and not rigid rules.
While modifiers appearing before a listing are often read to modify each element, that is not universally the case. See Dealertrack, Inc. v. Huber, No. CV 06-2335 AG (FMOx), 2008 WL 5792509, at *7 (C.D. Cal. Sept. 27, 2008) (holding SuperGuide presumption inapplicable in part because "`at least one of' [was] followed by a colon, and then a list separated by semicolons" which "implie[d] that the `at least one of' means `at least one of the elements listed below.'"); Antonin Scalia & Bryan A. Garner, Reading Law 150 (2012) (noting that this "series-qualifier canon" "[p]erhaps more than most of the other canons . . . is highly sensitive to context").
Courts engaged in statutory construction have not hesitated to read "and" as a disjunctive when warranted. See United States v. Fisk, 70 U.S. (3 Wall.) 445, 447 (1865) (to "ascertain the clear intention of the legislature, . . . courts are often compelled to construe `or' as meaning `and,' and again `and' as meaning `or.'"); Slodov v. United States, 436 U.S. 238, 246-47 (1978) (concluding that the word "and" was used disjunctively in 26 U.S.C. § 6672, which imposed a tax penalty on "[a]ny person required to collect, truthfully account for, and pay over any tax imposed by this title"); United States v. Pereira-Salmeron, 337 F.3d 1148, 1151 (9th Cir. 2003); 1A Norman J. Singer, Statutes and Statutory Construction § 21:14 (6th ed. 2003) ("There has been, however, so great laxity in the use of these terms that courts have generally said that the words are interchangeable and that one may be substituted for the other, if consistent with the legislative intent."); Webster's Third New International Dictionary 80 (2d ed. 2002) (defining "and" to include "reference to either or both of two alternatives <choose between him [and] me> esp. in legal language when also plainly intended to mean or <bequeathed to a person [and] her bodily issue> <property taxable for state [and] county purposes>").
In the asserted patents, "at least one of . . . and . . ." is plainly used in the manner proposed by PACT. To begin with, a number of claims are intelligible only under PACT's reading. Claim 4 of the '181 patent refers to "the second plurality of lines for at least one of reading data and writing data." By Altera's lights, this should either be read as "the second plurality of lines for reading data and writing data" — rendering "at least one of" entirely superfluous — or as "the second plurality of lines for at least one of reading data and at least one of writing data" — which is ungrammatical even for a patent. Constructions that read terms out of claims are to be avoided. See Digital-Vending Servs. Int'l, LLC v. University of Phoenix, Inc., 672 F.3d 1270, 1275 (Fed. Cir. 2012). And it is obvious that it makes no sense to speak of "at least one of reading data," since a gerund in this context is not a countable noun, so there cannot be "at least one of" it. By contrast, PACT's proposed reading, under which the term would mean "the second plurality of lines for reading data, writing data, or both" makes perfect sense of claim 4.
Similarly, claim 1 of the '106 patent says that "the plurality of the at least one individual lines, buses and subbuses at least one of combines multiple units and connects at least one of memories and peripherals." Under Altera's proposed construction, that should be read to mean, "connects at least one of memories and at least one of peripherals." Again, that is ungrammatical. What is evidently intended is "connects at least one of [a set composed of] memories and peripherals."
These examples from the claims of the bus system patents support the conclusion that the patentee intended "at least one of . . . and . . ." to be read as PACT suggests, and that that is how a person of ordinary skill in the art would understand them. The Court must presume that the other, more ambiguous, uses of the construction are to be read the same way. See Digital-Vending, 672 F.3d at 1275; Fin Control Sys. Pty, Ltd. v. OAM, Inc., 265 F.3d 1311, 1318 (Fed. Cir. 2001) ("[W]e begin with the presumption that the same terms appearing in different portions of the claims should be given the same meaning unless it is clear from the specification and prosecution history that the terms have different meanings at different portions of the claims.").
Altera's proposed construction would additionally conflict with the Summary of the Invention of the bus patents, which states that "[a]ny number of memories, peripherals, or other units can be connected to the bus system (e.g., for cascading)." See '181 patent, 1:56-58 (emphasis added).
Altera's proposed construction would also read a number of embodiments described in the patent specifications out of the asserted claims. Courts "normally do not interpret claim terms in a way that excludes embodiments disclosed in the specification." Oatey Co. v. IPS Corp., 514 F.3d 1271, 1276-77 (Fed. Cir. 2008) ("At leas[t] where claims can reasonably [be] interpreted to include a specific embodiment, it is incorrect to construe the claims to exclude that embodiment, absent probative evidence on the contrary."). Figures 10a-10f of the bus patents show various combinations of memories, peripherals, and other "units (DFPs, FPGAs, DPGAs, etc.)," but under Altera's proposed construction, the asserted claims of the '181 patent would require memories, peripherals, and an additional processing device, and therefore not cover those embodiments. By the same token, the specification of the '181 patent notes that "[i]n addition to cascading identical units (DFPs, FPGAs, DPGAs), memories and peripherals can also be connected as lower-level SLAVE units (SLAVE) to the bus system described here," '181 patent, 6:60-64 (emphasis added), implying, contrary to Altera's proposed construction, that the memories and peripherals are optional.
Every claim need not cover every embodiment described in the specification, see PSN Ill., LLC v. Ivoclar Vivadent, Inc., 525 F.3d 1159, 1166 (Fed. Cir. 2008), "especially . . . where . . . other unasserted claims in the parent patent cover the excluded embodiments," August Tech. Corp. v. Camtek Ltd., 655 F.3d 1278, 1285 (Fed. Cir. 2011). And claim 17 of the '077 patent (which is not asserted in this litigation) claims "[a] system configured to be connected to at least one of an external bus, a peripheral, or an external memory" (emphasis added), and would therefore likely cover Figures 10a-10f. But while it is theoretically possible that the patentee meant for none of the claims of the '181 patent to cover the embodiments in Figures 10a-10f, and chose to include those embodiments only within the scope of a few claims in a related patent, it defies common sense to hold that the patent should be read in such a contorted way.
Altera points out that a number of the asserted claims initially used the disjunctive "or" or "and/or," and were later amended to their present form. See, e.g., Dkt. Nos. 152-11 to 152-15. That at least raises the possibility that the patentee meant to narrow its claims to exclude Figures 10a-10f, and neglected to delete or modify the appropriate sections of the specification. But there is no evidence that the amendments that introduced the disputed claim language were meant to disclaim embodiments from the specification.
This is not to suggest that adopting PACT's proposal is without its own problems. For example, claim 1 of the '077 patent recites
'177 patent, 15:59-63, and later recites that the unit includes additional connections "with at least one of the DFP, the FPGA and the DPGA," id. 16:5-6. Under PACT's proposal, it would be possible for the unit recited earlier in the claim to include only an FPGA and the connections recited later in the claim to be with only "the DPGA," in which case "the DPGA" would have no antecedent basis. But the likely explanation for this textual difficulty is not, as Altera claims, that the patentee meant for the unit to include a DFP, FPGA, DPGA, and "a unit having a multi-dimensional programmable cell architecture," but rather that the patentee meant to say that the unit includes additional connections "with the at least one of a DFP, FPGA, and DPGA" — namely, the same one that was selected earlier in the claim.
Consequently, the Court adopts PACT's proposed construction of "at least one of . . . and. . . ."
D. "peripherals"/"peripheral devices"
Under the Court's supervision at the claim construction hearing, the parties came close to agreement on the proper construction of this term. Altera's main objection to PACT's proposed construction was that it might cover processors and reprogrammable devices, like FPGAs and DPGAs, see Altera's Claim Construction Brief at 5:16-22, Dkt. No. 150, despite the fact that the claims distinguish these devices from "peripherals," see, e.g., '077 patent, claim 1. But PACT agreed at the hearing that the term "peripheral" does not include DFPs or FPGAs, eliminating this potential dispute. See Hearing Tr. at 36:25-37:8.
There is a potential lack of clarity in Altera's own proposed construction, which is drawn almost verbatim from the sixth edition of the IEEE Standard Dictionary of Electrical and Electronic Terms, published in 1996. See Dkt. No. 152-4 at ALTR00286782. As Professor Wawrzynek said at the claim construction hearing, including the qualifier "not essential to the basic operation" could be inappropriate because peripherals are sometimes essential to the basic operation of the unit. See Hearing Tr. at 41:16-20. And that too assumes that the "basic operation" of the system can be determined with any level of reliability — a doubtful proposition.
At the hearing, PACT agreed to a compromise construction proposed by the Court, namely "an external device connected to the unit, such as a printer, a keyboard, a disk, a drive, or a converter," as long as a few chip-level devices were added to the list. See Hearing Tr. 38:20-40:7, 42:21-43:6. Altera objected to the proposed compromise construction on the basis that the reference to "the unit" was extraneous, as the patent claim itself describes what the peripheral is connected to. See Hearing Tr. 43:15-17, 43:24-44:2. Professor Wawrzynek pointed out that the term "graphic digital converter" is somewhat ambiguous. Wawrzynek Comments at 2. All of these points are well taken. The Court will delete the phrases "connected to the unit" and "graphic digital converter," but include in the list of examples a graphics processor unit (GPU) and a SCSI controller, the latter of which is expressly described in the specifications of the bus system patents as an "intelligent peripheral unit." See, e.g., '181 patent, 7:2-5.
The Court notes that the record at this point does not make entirely clear the parties' differences over the scope of the term "peripheral." To the extent later events cast the parties' dispute in sharper relief, the Court may find it necessary to clarify its construction. Pfizer, Inc. v. Teva Pharm., USA, Inc., 429 F.3d 1364, 1377 (Fed. Cir. 2005) ("district courts may engage in a rolling claim construction, in which the court revisits and alters its interpretation of the claim terms as its understanding of the technology evolves"); Network Commerce, Inc. v. Microsoft Corp., 422 F.3d 1353, 1358 n.4 (Fed. Cir. 2005) (approving a district court's clarification of its claim construction order on summary judgment). In the meantime, the Court construes "peripheral" and "peripheral device" to mean "an external device, such as a printer, keyboard, disk, graphics processing unit, or SCSI controller."
E. "bus system"
The parties' proposed constructions for "bus system" are nearly identical, save for one difference: Altera seeks to specify that the bus system must communicate information according to a "predefined" bus protocol, whereas PACT opposes the inclusion of the word "predefined." Altera initially proposed "fixed" instead of "predefined," but changed its mind in its opposition brief. Whichever word is used, Altera is clear that its construction is aimed at excluding bus systems that can be configured on the fly to use different bus protocols. Declaration of Majid Sarrafzadeh re Claim Construction ("Sarrafzadeh Decl.") ¶ 61, Dkt. No. 152.
The problem is that that exclusion does not appear in the intrinsic evidence. Altera points to places where the bus system control is described as "predefined" or "permanent," '181 patent at 1:43-56, but that says nothing about whether the bus system protocol can be reconfigured. Without such a limitation in the specification or prosecution history, "[i]t is improper for a court to add `extraneous' limitations to a claim, that is, limitations added `wholly apart from any need to interpret what the patentee meant by particular words or phrases in the claim.'" Hoganas AB v. Dresser Indus., Inc., 9 F.3d 948, 950 (Fed. Cir. 1993); see also Amgen Inc. v. Hoechst Marion Roussel, Inc., 314 F.3d 1313, 1325 (Fed. Cir. 2003) (holding that claims should not be limited to only exogenous DNA where claims included no such limitation).
Implicitly acknowledging that PACT's proposed construction does not actually contradict anything in the specification, Altera argues that its scope is not "supported or enabled" by the written description. See Altera's Claim Construction Brief at 6:23-24. But "[e]nablement concerns do not justify departing from the plain and ordinary meaning" where the meaning of a term is clear, because claim terms "cannot be rewritten by the courts to save their validity." See Hill-Rom Servs., Inc. v. Stryker Corp., 755 F.3d 1367, 1374 (Fed. Cir.), cert. denied, 135 S.Ct. 719 (2014); Phillips, 415 F.3d at 1327-28 (noting that the maxim of interpreting claims to preserve their validity has "not [been] applied . . . broadly" and is "a doctrine of limited utility").
In any event, the evidence that Altera's construction is necessary to avoid enablement issues is, at this stage, ambiguous and inconclusive. The bus system patents do not discuss how to create an embodiment that switches between bus protocols.
The opinion of Altera's expert, Professor Sarrafzadeh, that a person of ordinary skill in the art would not be able to design an embodiment of the invention that switches bus protocols is too conclusory to meet that standard. See Sarrafzadeh Decl. ¶ 65. It is also contradicted by the opinion of the court-appointed expert, Professor Wawrzynek, who stated at the claim construction hearing that a person of ordinary skill in the field at the time of the invention would be able to implement a bus system control that switches between protocols — for example, by switching between multiple state machines. Hearing Tr. at 50:19-51:9. This conflicting evidence regarding potential enablement problems is a dubious ground on which to import a limitation with little intrinsic support into the claims.
In declining to narrow the construction of "bus system," the Court does not prejudge any potential future claim by Altera that the claims are invalid for lack of enablement. The Federal Circuit has previously construed terms to avoid narrowing limitations, and then found the full scope of the term unenabled. Liebel-Flarsheim Co. v. Medrad, Inc., 481 F.3d 1371, 1378-79 (Fed. Cir. 2007) (finding full scope of claim term not enabled after earlier appeal construing term broadly). The Court expresses no opinion based on the limited record currently before it as to the validity of the claims that contain the "bus system" term.
Accordingly, the Court construes "bus system" to mean "a system used to communicate information according to a bus protocol."
The main dispute with respect to the term "unit" is over the word "complete": Altera wants to include it, while PACT thinks it unnecessary. Altera's main selling point for its proposed construction is that it attempts to delineate the contours of the unit, by including only those portions of the device that are necessary to perform the stated function of any given "unit." PACT, on the other hand, does not take issue with the coverage of Altera's proposed claim construction so much as its lack of clarity, arguing that the inclusion of the word "complete" simply pushes the question back to whether a given piece of circuitry is, in fact, complete.
Although Altera argues that its proposed construction is not inconsistent with the intrinsic evidence, it does not point to anything in the specification or prosecution history that compels its proposed construction. Instead, it cites two dictionary definitions in support. See IEEE Standard Dictionary of Electrical and Electronics Terms 1162 (6th ed. 1996) (defining "unit" as, among other things, "[a] major building block for a set or system, consisting of a combination of basic parts, subassemblies, and assemblies packaged together as a physically independent entity."), Dkt. No. 152-4; Webster's New World Dictionary of Computer Terms 598 (1994) (defining "unit" as "[a]ny device having a special function, such as the arithmetic-logic unit, central processing unit, or disk unit."), Dkt. No. 152-9. Two things are immediately apparent from these definitions. The first is that neither requires that the unit be "complete," or that it be limited to include the circuitry necessary to performing a certain function and only that circuitry. The second is that neither definition adds much to a lay understanding of the term. Under those circumstances, there is no need to construe the claim term at this point. See O2 Micro Int'l Ltd. v. Beyond Innovation Tech. Co., 521 F.3d 1351, 1362 (Fed. Cir. 2008) ("district courts are not (and should not be) required to construe every limitation present in a patent's asserted claims."); Biotec Biologische Naturver. GmbH v. Biocorp, Inc., 249 F.3d 1341, 1349 (Fed. Cir. 2001) ("the meaning of `melting' does not appear to have required `construction,' or to depart from its ordinary meaning"); Bd. of Trustees of Leland Stanford Jr. Univ. v. Roche Molecular Sys., Inc., 528 F.Supp.2d 967, 976 (N.D. Cal. 2007) ("The terms `therapeutically effective' or `therapeutically ineffective' are commonplace—a juror can easily use these terms in her infringement fact-finding without further direction from the court. These terms do not need to be construed because they are neither unfamiliar to the jury, confusing to the jury, nor affected by the specification or prosecution history."). Expert discovery and motion practice may later reveal a "fundamental dispute" between the parties regarding claim scope, obliging the Court to clarify the meaning of the term, see O2 Micro, 521 F.3d at 1362, but so far no such disagreement is apparent. As previously noted, PACT argues that requiring a unit to be "complete" introduced confusion, not that it impermissibly expands or contracts the scope of the claims.
Consequently, the Court leaves "unit" unconstrued.
G. "interface"/"interface unit"
The parties agree that "interface" and "interface unit" should have the same construction, and at the hearing, came to agreement as to what that construction should be. Both parties agreed on the record to a compromise construction proposed by the Court: "a unit providing permanent implementation of a bus system control without necessarily requiring any influence by a programmer for communicating information across a shared boundary." Hearing Tr. at 68:22-70:4. As stated at the hearing, the Court adopts that compromise construction of "interface" and "interface unit." The parties' agreement does not waive any positions they may wish to take in the future regarding infringement and invalidity under this construction. See id. at 69:17-22, 74:23-75:15.
H. "interface elements configured to share the external bus"
The parties' proposed constructions boil down this question: Does "to share" the external bus mean "to permit access to," as PACT argues, or "to make competing demands for," as Altera would have it? Under PACT's proposed construction, it is enough if the interface elements can access the external bus. Under Altera's, they must actually compete for use of the external bus.
Turning first to the specification, the controller disclosed by the '077 patent is described as having two modes: an "active mode" in which the controller "controls the internal bus (I-BUS) and the external bus (E-BUS)" and a "passive mode" in which the controller "controls only the internal bus (I-BUS)" but "[t]he E-BUS is controlled by another external unit," with the controller simply "react[ing] in this mode to the requirements of the external [unit]." '077 patent, 4:66-5:10. In passive mode, then, the controller would not "share" the external bus under Altera's proposed construction, and would therefore not be covered by claims 14 and 15. Altera's proposed construction, then, would read out certain disclosed embodiments of the '077 patent. Indeed, Altera's expert conceded that under Altera's construction, "none of the embodiments" in the '181 patent would be covered by claim 14 of the '077 patent, including specifically Figures 9A and 9B of the '181 patent. See May 8, 2015, Deposition of Majid Sarrafzadeh 170:10-21, Dkt. No. 160-2. Most of these embodiments of the '181 patent (including Figures 9A and 9B) also appear in the '077 patent, meaning that Altera's proposed construction would narrow claim 14 to exclude disclosed embodiments.
Altera says that "where the patent describes multiple embodiments, every claim does not need to cover every embodiment," and suggests that claims 14 and 15 of the '077 patent were merely meant to cover embodiments disclosed in Figures 13-17, which were newly added during prosecution of the '077 patent and not included in the earlier related '181 patent. Pacing Techs., LLC v. Garmin Int'l, Inc., 778 F.3d 1021, 1026 (Fed. Cir. 2015); PSN, 525 F.3d at 1166 ("[C]ourts must recognize that disclosed embodiments may be within the scope of other allowed but unasserted claims."). But the principle that not every claim need cover every embodiment is usually applied in cases where "the plain language of a limitation of the claim does not appear to cover that embodiment," Pacing Techs, 778 F.3d at 1026, and "the court's construction is supported by the intrinsic evidence," August Techs. Corp. v. Camtek, Ltd., 655 F.3d 1278, 1285 (Fed. Cir. 2011) (quoting TIP Sys., LLC v. Phillips & Brooks/Gladwin, Inc., 529 F.3d 1364, 1373 (Fed. Cir. 2008)). Here, by contrast, Altera argues on the basis of extrinsic evidence (namely, the expert declaration of Prof. Sarrafzadeh, see Sarrafzadeh Decl. ¶ 90) that the plain and ordinary meaning of "sharing" an external bus is to make competing demands on it. But that is only one of multiple meanings of the word "share"; it is perfectly grammatical to speak of sharing read access to a bus, but not sharing the ability to write to the bus. See, e.g., IBM Dictionary of Computing 619 (George McDaniel ed., 10th ed. 1993) (defining "shared" as "[p]ertaining to the availability of a resource for more than one use at the same time").
Claim differentiation principles also support the conclusion that "sharing" in the context of claims 14 and 15 does not require making competing demands on the bus. "In the most specific sense, claim differentiation refers to the presumption that an independent claim should not be construed as requiring a limitation added by a dependent claim." Curtiss-Wright Flow Control Corp. v. Velan, Inc., 438 F.3d 1374, 1380 (Fed. Cir. 2006) (citing Nazomi Comm'ns, Inc. v. Arm Holdings, PLC, 403 F.3d 1364, 1370 (Fed. Cir. 2005)). Claim 15 depends from independent claim 14, but adds the limitation, "an arbiter connected to each of the plurality of interface elements." '077 patent, claim 15. The doctrine of claim differentiation implies that claim 14 does not require an "arbiter," which the parties agree means "circuitry that resolves competing demands for access to a common resource." See Joint Claim Construction Statement, Ex. B at 2, Dkt. No. 114-2. It follows that it must be possible to have "interface elements configured to share the external bus" without having a means for resolving competing demands to the bus — which is possible only if the system is willing to live with collisions on the bus, or if the interface elements can share the external bus without making competing demands on it. The former possibility is foreclosed by the specification's admonition that "the [primary logic unit] must be sure that there are no bus collisions on the [external bus]." See '077 patent, 5:53-56. Consequently, the interface elements can "share" the external bus without making competing demands on it, contrary to Altera's proposed construction.
The Court adopts PACT's proposed construction of "interface elements configured to share the external bus," namely "circuitry providing permanent implementation of a bus system control for communicating information across a shared boundary configured to permit access to the external bus."
While PACT would prefer leaving this term unconstrued, Altera proposes that it be construed to mean "to exercise authoritative or dominating influence over something." Altera's construction is drawn entirely from the definition of "control" found in a single general-purpose dictionary. See The American Heritage Dictionary of the English Language 410 (3d ed. 1996) (giving one definition of "control" as "[t]o exercise authoritative or dominating influence over; direct").
Altera's proposed construction well illustrates the danger the Federal Circuit warned of when it observed that "[i]n the course of litigation, each party will naturally choose the pieces of extrinsic evidence most favorable to its cause. . . ." Phillips, 415 F.3d at 1318. The American Heritage Dictionary may insist that to "control" is necessarily to "exercise authoritative or dominating influence," but other dictionaries reveal deeper shades of meaning. See Dictionary of Computing 105 (Oxford Univ. Press 4th ed. 1996) (defining "control circuitry" as "[e]lectric circuits within a computer or peripheral that regulate its operation"); Microsoft Computer Dictionary 110 (4th ed. 1999) (defining "control" to mean "[m]anagement of a computer and its processing abilities so as to maintain order as tasks and activities are carried out"); 3 Oxford English Dictionary 853 (2d ed. 1989) (defining "control" as "[t]o exercise restraint or direction upon the free action of; to hold sway over, exercise power or authority over; to dominate, command"). There is no evidence in the patent or file history that the patentee intended to avoid these nuances when it used the term "control," and there is no reason why a lay jury would need instruction on the meaning of a common English word used in its ordinary sense. See ProconGPS, Inc. v. Skypatrol, LLC, No. C 11-03975 SI, 2012 WL 3276977, at *4-5 (N.D. Cal. Aug. 9, 2012) (finding "the term `loan' does not require construction" because it was "`as common a word as one could possibly encounter in a patent case'"); Netflix, Inc. v. Blockbuster, Inc., 477 F.Supp.2d 1063, 1068 (N.D. Cal. 2007) ("As [the terms `movie' and `item'] are commonly-understood English words, they need no clarification."). The Court consequently leaves "control," "controls," and "controlling" unconstrued.
J. "compiler"/"configuration unit"
The parties agree that "compiler" and "configuration unit" are synonymous. The point of difference between the parties is that, under Altera's proposed construction, the "compiler" and "configuration unit" must be "[e]mbodied by a microcontroller specifically designed for this purpose." See Final Joint Claim Construction Statement at 12. PACT claims that including this limitation is inconsistent with the specifications of the '242 and '869 patents, which describe the compiler as being "based on a transputer 31, i.e., a processor with a microcoded set of instructions." See '242 patent at 6:20-21; '869 patent at 6:27-28. If a microcontroller is already specifically designed to be a compiler or configuration unit, PACT's argument goes, it does not need to be programmed in microcode after the fact to perform its function. Altera, on the other hand, says that the language from the specification only excludes using an ASIC — an application-specific integrated circuit — as the compiler, and that its construction does not require the compiler to be an ASIC. See Villasenor Decl. ¶ 64.
Altera's main argument is that the patent specifications do not enable a "compiler" that is not specially designed for its purpose, so the compiler must either be "[e]mbodied by a microcontroller specifically designed for [its] purpose" or the patent is invalid for lack of enablement. See Altera's Claim Construction Brief at 13:1-17. As with the "bus system" term, however, it is generally inappropriate to limit the scope of a claim based on enablement concerns, especially where, as here, the expert opinion offered by Altera does not attempt to demonstrate that the asserted claims are not enabled under the relevant standard, namely "that a person of ordinary skill in the art would not be able to practice the claimed invention without `undue experimentation.'" See Alcon Research Ltd. v. Barr Labs, Inc., 745 F.3d 1180, 1188 (Fed. Cir. 2014); Villasenor Decl. ¶¶ 62-65.
Altera's expert also suggests that it would be "evident to one of ordinary skill in the art" that "compiler" and "configuration unit" refer to a microcontroller. See Villasenor Decl. ¶¶ 64-65. But all he cites in support is the definition of a "primary logic unit" or "PLU" from the '525 patent — a patent from a different family — at column 17 lines 52 to 54, and another court's construction of "compiler" in a prior litigation by PACT against Altera's competitor Xilinx,
Without this evidence, the Court is left only with Altera's expert's bare opinion that a person of ordinary skill in the art would understand that a compiler or configuration unit are embodied in a microcontroller. This conclusion, however, is contradicted by the Court-appointed expert's opinion that the "embodied by a microcontroller" limitation is inappropriate, because the compiler or configuration unit could be embodied in dedicated logic. See Wawrzynek Comments at 5. Such ambiguous extrinsic evidence is not a proper basis on which to include the limitation Altera proposes. See Texas Digital Sys. v. Telegenix, Inc., 308 F.3d 1193, 1212 (Fed. Cir. 2002) (holding that expert testimony "may not be used to vary or contradict the claim language").
Accordingly, the Court will not include the requirement that the "compiler" and "configuration unit" be "[e]mbodied by a microcontroller specifically designed for this purpose," and instead construes both terms to mean "unit for configuring and reconfiguring a PAE or logic cell."
K. "configuration memory"/"configuration signal(s)"/"configuration interface"
Once again, the parties' proposed constructions are identical, apart from one additional limitation proposed by Altera, namely that the configuration bits stored in the memory must represent "a valid setting . . . so that an operable unit is obtained." See Joint Claim Construction Statement at 14.
For the "configuration memory" term,
Altera's argument assumes that claim 14 actually requires the configuration memory to contain configuration signals. But it simply requires that the configuration memory be "adapted to store the plurality of configuration signals," '869 patent, 14:3-4 (emphasis added), which is perfectly consistent with the memory not containing configuration signals at any given point in time.
Additionally, for all three terms, there is no reason to require that the configuration memory only contain configuration signals that represent a "valid setting" and result in an "operable unit." Altera and its expert point to no intrinsic evidence to support these limitations, apart from simply claiming that they are implicit in the concept of "configuration" itself. See Villasenor Decl. ¶ 74. But there is no reason why there cannot be a configuration that turns out to be inoperable (for example, during testing), and if there is, the "configuration signal" does not cease to be one thereby. Without such intrinsic evidence, or non-conclusory extrinsic evidence, it is inappropriate to add limitations to the claims under the guise of construction. See Amgen, 314 F.3d at 1325.
Altera also points out that the limitation it is proposing was successfully championed by PACT in the Xilinx litigation. But the point of the exercise of claim construction is to "to clarify and when necessary to explain what the patentee covered by the claims," not to grade the parties on consistency. U.S. Surgical Corp. v. Ethicon, Inc., 103 F.3d 1554, 1568 (Fed. Cir. 1997). Altera made clear at the claim construction hearing that it does not contend that the claim construction PACT advanced in the Xilinx litigation is binding, apart from its being (in Altera's view) correct. See Hearing Tr. 70:15-72:21. In the Court's view, it is not correct.
Finally, Altera argues that PACT's proposed construction would allow the configuration unit to contain bits that have nothing to do with configuration, like control signals. Altera's concern that PACT's proposed construction would encompass a configuration memory that stored signals that have nothing to do with configuration appears to be adequately addressed by the fact that PACT's proposed construction requires that the signal "represent a setting for the element to be configured." Non-configuration-related signals, like power and clocking, presumably do not represent settings for the configured element. To the extent that PACT does, in fact, assert that these claim terms cover control signals, the Court may revise the claim constructions. Network Commerce, 422 F.3d at 1358 n.4.
Meanwhile, the Court construes these three terms as stated in the table above.
L. "configuration"/"configur[-e/-ing/-ed]"/"configurable"/"reconfigur[-e/-es/-ing/-ed]"/"reconfigurable"/"reconfigurable cells dynamically reconfigur[-e/-ing]"/"dynamically reconfigurable cells"/"configurable [at/during] runtime"
This formidable list of claim terms turns out to involve only a single dispute — and even that dispute proves illusory. In each of its proposed constructions for these terms, Altera wishes to include language emphasize that the process of configuring or reconfiguring involves setting the "configuration memory bits that dictate" the function and interconnection of a logic unit, FPGA cell, logic cell, or PAE. PACT wants to leave this language out.
Altera and PACT appear to agree that configuration will involve setting bits in memory. See PACT's Claim Construction Brief at 18:9-13, Dkt. No. 143; Altera's Claim Construction Brief at 18:3-8. The parties also appear to agree that no reference to "configuration memory bits" is actually found in the intrinsic record. See Villasenor Decl. ¶¶ 90-96 (quoting intrinsic evidence, none of which refers to "configuration memory bits"). Altera suggests that its proposed language should be added to the construction of this claim element to ensure that the jury and PACT are aware of "the context of the alleged inventions claimed in the patents-in-suit." Altera's Claim Construction Brief at 18:11-13. But that context is better provided through explanations from the various testifying experts — both those retained by the parties and the court-appointed expert — than through the inclusion of a claim limitation that is itself likely obscure to a lay jury.
Accordingly, the Court construes the terms in this group as stated in the table above.
A. "the unit"
In the patent milieu, "the" is often used with a specialized meaning: "the fruit," like "said fruit," refers back to "a fruit" recited somewhere earlier in the claim. See Tuna Processors, Inc. v. Hawaii Int'l Seafood, Inc., 327 F. App'x 204, 210 (Fed. Cir. 2009) (non-precedential). But what if there are two instances of "a fruit" in the claim? That is what Altera says is the problem with claim 1 of the '077 patent, which recites:
A bus system, comprising:
'077 patent, claim 1 (emphasis added). Does "the unit" refer back to the first instance of "a unit," underlined above? Or the second instance, in italics? Or the double-underlined phrase "multiple units"?
The United States Patent and Trademark Office's Manual of Patent Examining Procedure suggests that claims might be rendered indefinite, and therefore invalid, by ambiguous antecedents:
MPEP 2173.05(e). The manual also states, however, that antecedent basis problems do not automatically invalidate a claim: "Obviously, however, the failure to provide explicit antecedent basis for terms does not always render a claim indefinite. If the scope of a claim would be reasonably ascertainable by those skilled in the art, then the claim is not indefinite." Id.
The "MPEP does not have the force of law, and is only entitled to judicial notice as the PTO's official interpretation of statutes and regulations with which it is not in conflict." Belkin Int'l v. Kappos, 696 F.3d 1379, 1384 (Fed. Cir. 2012) (citing Molins PLC v. Textron, Inc., 48 F.3d 1172, 1180 n.10 (Fed. Cir. 1995)). And "indefiniteness rejections by the USPTO arise in a different posture from that of indefiniteness challenges to an issued patent." In re Packard, 751 F.3d 1307, 1312 (Fed. Cir. 2014). In particular, due to "the important role that the USPTO, through its examination process, plays in ensuring the quality of patents," the examiner may issue an indefiniteness rejection to clarify a proposed claim, even if a court examining the claim post-issuance would not necessarily find it invalid. Id. at 1324 (Plager, J., concurring). The MPEP is meant for patent examiners to use in fulfilling that role. The Court, on the other hand, applies the standard indefiniteness test, and asks whether the "claims, read in light of the specification delineating the patent, and the prosecution history, fail to inform, with reasonable certainty, those skilled in the art about the scope of the invention." Nautilus, 134 S. Ct. at 2124.
Altera and PACT have submitted competing expert declarations on what the answer is. Altera's expert, Professor Majid Sarrafzadeh, opines that the claim does not inform a person of ordinary skill with ordinary certainty which prior unit is being referred to, while PACT's expert, Dr. Marwan Hassoun, opines that it does. See Declaration of Majid Sarrafzadeh re Invalidity ("Sarrafzadeh Invalidity Decl.") ¶¶ 13-26, Dkt. No. 142;
According to PACT and its expert, "the unit" refers back to the first (underlined) instance of "a unit." They contend that the second (italicized) instance, "a unit having a multi-dimensional programmable cell architecture," is a catch-all term for all multi-dimensional programmable cell architectures, of which DFPs, FPGAs, and DPGAs are examples. Hassoun Decl. ¶ 47. PACT's expert suggests that the use of this unwieldy term is "necessary because there is no well-known term or acronym in the art encompassing all the different possible multi-dimensional programmable cell architectures." Id.
The Court agrees with PACT that the claim conveys, with reasonable certainty, that "the unit" refers to the first instance of "a unit." The Court has construed "a unit including at least one of a data flow processor (DFP), a field programmable gate array (FGPA), a dynamically programmable gate array (DPGA), and a unit having a multi-dimensional programmable cell architecture" not to require a DFP and an FPGA and a DPGA and "a unit having a multi-dimensional programmable cell architecture." Given this construction, it is possible that an embodiment of the claim would not include "a unit having a multi-dimensional programmable cell architecture." But if that is what "the unit" refers back to, it would, in some cases, have no antecedent at all.
Similarly, "the unit" cannot refer back to "multiple units," because "the unit" implies a single unit. See Howmedica Osteonics Corp. v. Wright Medical Tech., Inc., 540 F.3d 1337, 1344 (Fed. Cir. 2008) (holding that a reference to "the condylar element" meant "only one condylar element").
Consequently, the Court finds that the use of "the unit" in claim 1 of the '077 patent does not render the claim indefinite.
Claim 13 of the '869 patent recites:
'869 patent, claim 13 (emphasis added). Claim 14 depends from claim 13, and therefore includes the same limitation.
The parties agree that "haltered" makes no sense in the context of this claim. See Altera's Motion for Summary Judgment at 8:24-25, Dkt. No. 140; PACT's Opposition ("PACT's Opp.") at 7:20-21, Dkt. No. 153. Altera argues that it is unclear whether the intended word was "halted," or "altered," while PACT argues that "haltered" was a typographical error for "halted."
Although not cited by either party, the Federal Circuit has addressed the circumstances in which a district court can correct a typographical error in a patent on claim construction, even when no certificate of correction has been issued. It has held that:
Novo Indus., L.P. v. Micro Molds Corp., 350 F.3d 1348, 1354 (Fed. Cir. 2003). Under this standard, the power of the district court to correct a patent is more limited than the Patent and Trademark Office's ability to do so by issuing certificates of correction under 35 U.S.C. §§ 254-55. See id. at 1356. That is necessarily the case, for otherwise there would be no point to the non-retroactivity provisions Congress included in the 1952 statute authorizing certificates of correction, which provide that certificates of correction are only effective for suits brought after the certificate is issued. See id. at 1355-56. A patentee faced with a situation where the statute did not authorize a certificate of correction could just ask a court for the equivalent, rendering the provisions toothless.
Among the differences between the PTO's and the district court's power to correct patents is the fact that the district court can only correct "obvious" errors that are "evident from the face of the patent itself," while the PTO can review the entire "intrinsic record," including, for example, the prosecution history. See id. at 1356-57. In other words, for the district court to correct a patent, the correction must be obvious from looking just at the claims and the specification, without needing to turn to the prosecution history — though the district court must still consider whether the prosecution history tells against the proposed correction. See H-W Tech., L.C. v. Overstock.com, Inc., 758 F.3d 1329, 1334 (Fed. Cir. 2014) ("[T]his court has already deemed evidence of error in the prosecution history alone insufficient to allow the district court to correct the error."); Group One, Ltd. v. Hallmark Cards, Inc., 407 F.3d 1297, 1303 (Fed. Cir. 2005) ("The prosecution history discloses that the missing language was required to be added by the examiner as a condition for issuance, but one cannot discern what language is missing simply by reading the patent. The district court does not have authority to correct the patent in such circumstances.").
Yet the primary evidence PACT points to in arguing that "haltered" should be read as "halted" is the prosecution history of the '869 patent and the specification of the related '242 patent, not the specification and claims of the '869 patent itself. See PACT's Opp. at 8:2-9:1; Hassoun Decl. ¶¶ 58-59. The original application of the '869 patent used the word "halted" where "haltered" is now used. See Dkt. No. 153-5 at PACT00015659. In addition, the '242 patent states:
See '242 patent, 8:56-60 (emphasis added). The '869 patent, in an identical portion of the specification, uses "baited." See '869 patent, 9:1-5. But neither of these are drawn from the face of the '869 patent itself, so the Court cannot use them to correct the claims.
PACT's expert also opines that a person of ordinary skill in the art at the time of the invention of the '869 patent "would understand with at least reasonable certainty" that claim 13 should read "halted" instead of "haltered." Hassoun Decl. ¶ 56. But the only evidence from the face of the patent that he cites is from the '869 patent's Abstract, which states:
'869 patent, Abstract (emphasis added). According to PACT and its expert, the whole point of the invention is to allow reconfiguration of certain cells or groups of cells without "halting" or "deactivating" neighboring cells. But it is not enough that PACT's interpretation simply be the more likely one; rather, it must be "obvious" and "not subject to reasonable debate." Novo, 350 F.3d at 1356-57. As Altera points out, a number of portions of the '869 patent's specification state that the neighboring cells must not be "affected" — which is closer in meaning to "altered" than to "halted" or "deactivated":
'869 patent, 2:22-26 (emphasis added).
Id. at 2:50-55 (emphasis added); see also id. at 6:63-68. Whether "haltered" in claim 13 of the '869 patent means "halted" or "altered" is therefore not beyond the realm of reasonable debate, at least if looks only to the claims and specification. Where, as here, "the claim language `might mean several different things and no informed and confident choice is available among the contending definitions,'" the claim is invalid for indefiniteness. See Interval Licensing LLC v. AOL, Inc., 766 F.3d 1364, 1371 (Fed. Cir. 2014) (quoting Nautilus, 134 S. Ct. at 2130) (internal quotations omitted). It rightly is not disputed that "halted" and "altered" mean different things.
As a result, the Court finds that claims 13 and 14 of the '869 patent are invalid.
The Court sets a case management conference to discuss further dates on September 30, 2015. The parties are ordered to file a joint case management statement no less than seven calendar days before then.
OfficeMax, Inc. v. United States, 428 F.3d 583, 600 (6th Cir. 2005) (Rogers, J., dissenting). Altera would presumably think the first guest would be happy with a glass of water.