Network Appliance, Inc. v. Bluearc Corp.

374 F. Supp. 2d 825, 2005 U.S. Dist. LEXIS 16741, 2005 WL 1530098

• Court: District Court, N.D. California
• Decided: June 27, 2005
• Docket: C 03-5665 MHP
• Status: Published

Opinion

MEMORANDUM & ORDER

PATEL, District Judge.

Re: Cross-Motions for Summary Judgment of Validity and Invalidity

Plaintiff Network Appliance filed this action in the United States District Court for the District of Delaware, alleging infringement of three patents that relate to network file server architecture and operating system software. On defendant’s motion, the District of Delaware ordered the action transferred to this court on December 16, 2003. Now before the court are the parties’ cross-motions for summary judgment of validity and invalidity of United States Patent Nos. 5,802,366 (“ ’366 Patent”), 5,931,918 (“’918 Patent”), and 6,065,037 (“’037 Patent”). Having fully considered the parties’ arguments and submissions and for the reasons set forth below, the court enters the following memorandum and order.

BACKGROUND

I. The Patents in Suit and the Prior Art

The patents that are the subject of this infringement action relate to computer architectures and storage systems that were developed by Auspex Systems, Inc. in the late 1980s. 1 The ’366 Patent, issued on September 1, 1998, and the ’918 Patent, issued on August 3, 1999, disclose substantially identical multiprocessor file server designs (“architectures”) employing separate processors to perform file system control and non-file system control tasks. The ’037 Patent, issued on May 16, 2000, describes a computer operating system that allows for the implementation of the file server designs disclosed by the ’366 and ’918 Patents. Each of these patents claims priority based on a parent application filed on September 8, 1989, which is the undisputed date of invention (i.e., “the critical date”) of the patents.

The court has set forth a detailed description of the patented inventions in a number of previous orders and need not repeat that description here. Rather, the instant motions for validity and invalidity direct the court’s attention to the contents of the prior art on the effective filing date of the patents in suit. The following discussion summarizes the pertinent prior art references.

A. The V-System

The first prior art reference cited by defendant describes the “V-System” com puter operating system developed at Stanford University’s Computer Science Laboratory. See Kent. Decl., Exh. C, David R. Cheriton & Keith A. Lantz, V-System 6.0 Reference Manual (1986) (hereinafter “V-System Manual”). The V-System Manual describes a computer network in which individual workstations are treated as components of a “distributed” operating system. Id. at 1-1 to 1-3. To implement this architecture, the V-System relies on a “distributed kernel,” which, roughly speaking, consists of a stripped-down version of a complete Unix operating system that manages certain “low-level” functions and provides for communication among the various workstations that make up the operating system. Id. at 1-3. These workstations are then assigned specific tasks such as handling “exceptions” (i.e., operating system errors) and managing data storage devices. See id. at 31-2 to 31-5. The main purpose of this architecture is to “partition[] functions as far as practical into separate servers,” thereby keeping the kernel and servers “reasonably small” in size and independent of each other in operation. Id. at 31-8.

The V-System’s software is also described in an article by David R. Cheriton, entitled The V Kernel: A Software Base for Distributed Systems, 1 IEEE Software 19 (1984). Wales Decl., Exh. B (hereinafter “Cheriton”). Cheriton provides a more detailed description of the V-System’s distributed kernel, focusing on the “interpro-cess communications” component of the kernel that allows for the passing of messages among the server’s constituent workstations. Id. at 21-25.

B. The Epoch Server

The next prior art reference relevant to the instant motions is the Epoch-1 Infinite Storage Server (“the Epoch Server”), which is described in an article that appeared in Computer Design on November 15, 1988, Kent Decl., Exh. D (hereinafter “Lieberman”), as well as in an article published in Digital Review on May 22, 1989, id., Exh. E (hereinafter “Haskin”). These articles describe “a general-purpose file server that implements a hierarchy of solid state, magnetic and optical memory,” the primary benefit of such a design being the relative ease with which archiving and backup of system data are achieved. Lieberman at 36; see also Haskin at 31. However, for purposes of the instant motion, the most significant aspect of the Epoch Server is the fact that it employs a multiprocessor architecture comprised of two 25-MHz Motorola 60820 processors, the first processing unit being a “front-end I/O [i.e., input-output] processor” and the second unit serving as a file processor. Haskin at 32. The front-end processor runs a “low-overhead” — that is to say, relatively simple and fast — operating system (known as “VRTX”) and handles all input-output operations for the server. Id.; McKusick Decl. in Support of Def.’s Mem. P. & A. ¶ 42 (hereinafter “First McKusick Decl.”). The file processor employs the “enhanced BSD Unix” operating system and is “primarily concerned with handling ... NFS requests from other Unix hosts,” although it is also capable of performing “other Unix housekeeping chores.” Has-kin at 32. In addition, the Haskin reference discloses a “shared memory” used to pass data from one processor to the other. Id.

C. Sandberg

The next reference upon which defendant’s invalidity contentions rely is an April 1986 article by Russell Sandberg, The Sun Network Filesystem: Design, Implementation, and Experience, Proceedings of the 1986 European Unix User’s Conference (1986). Kent Decl., Exh. I (hereinafter “Sandberg”). In that paper, Sandberg discusses the design and implementation of the Sun Network Files- ystem (“NFS”) protocol, a widely adopted standard for performing file system-related operations in a Unix-based networked computing environment. See generally id. After describing the details of the NFS protocol, Sandberg briefly turns to his experiences with “porting” NFS to various hardware and software systems. Id. at 10410. Of particular relevance here is Sandberg’s discussion of the implementation of NFS on a VAX 750 processor running the System V.2 operating system (a version of Unix). See id. The article describes the use of an “Excelan” networking board for the purpose of handling the Ethernet, IP, and UDP layers (i.e., the network and transport layers) of NFS requests. Id. 2 A number of other ports of NFS are also described. See id.

D.Figure 1 of the ’918 Patent

The final reference that defendant’s motions identify as anticipating the patented inventions is the ’918 Patent’s own depiction of prior art file servers. In Figure 1 of the patent, the inventors disclose a block.diagram of the prior art file server architecture. ’918 Patent, 4:57-58 & fig. 1. The patent describes the prior art server as having a host CPU connected to an Ethernet, a large memory array, and a number of peripheral devices, including a disk controller managing one or two disk drives, a tape controller connected to a tape drive, and “possibly” a network controller connected to a second Ethernet network. Id. at 5:14-24. The patent’s written description also describes how the disclosed prior art server responds to NFS requests from a server located on the second Ethernet. Id. at 7:14-39.

E. Coulouris

In addition to the allegedly anticipatory references cited in defendant’s motions, several additional prior art references are identified by the parties as prior art relevant to the issues of novelty and nonobvi-ousness. The first of these references is a textbook written by George F. Coulouris and Jean Dollimore and published in 1988. Fink Deck, Exh. D, George F. Coulouris & Jean Dollimore, Distributed Systems: Concepts and Designs (1988) (hereinafter “Coulouris”). The authors provide a general overview of distributed file server architectures of the type disclosed in the V-System Manual. See generally Coulouris at 25-32. Among the topics discussed is the implementation of the Sun NFS files-ystem on a distributed network. Id. at 264-69.

F. Palowski

The next reference identified as pertinent prior art is a 1989 article published by Brian Palowski and numerous co-authors. Fink Decl., Exh. H, Brian Palowski et al., Network Computing in the Unix and IBM Mainframe Environment, Uni-Forum Conference Proceedings, at 287 (1989) (hereinafter “Palowski”). The Pa-lowski reference teaches the implementation of the Sun NFS filesystem in an IBM mainframe computing environment, disclosing a file server architecture comprised of a Sun “front-end processor” and an MVS mainframe computer and configured to respond to NFS requests from client computers on a network. Id. at 294-95 & fig. 3. The front-end processor handles lower-level protocol processing for incoming file system requests, with the task of processing higher-level protocol layers (including the NFS application layer) being assigned to the mainframe. Id. According to the authors, the primary advantage of this design lies in the segregation of data management related-tasks from “user interface and data manipulation functions,” the former being performed by the mainframe computer and the latter being “offload[ed]” onto other computers on the network. Id. at 291.

G. Rosen

The final relevant prior art reference is an article presented at the Advanced Computing Systems Association (“USENIX”) Summer Conference in 1986. Fink Deck, Exh. I, Mordecai B. Rosen et al., NFS Portability, Proceedings of the USENIX 1986 Summer Conference, at 289 (1986) (hereinafter “Rosen”). Rosen teaches the implementation of Sun NFS in a multiprocessor computing environment based on the System V.2 Unix operating system. Id. at 299-300, 302. The authors characterize their experience with the System V.2 port as “demonstrat[ing] the feasability of re-hosting ... NFS to new environments.” Id. at 300.

II. Procedural History

On July 15, 2003, plaintiff filed an action in the United States District Court for the District of Delaware, alleging that defendant had infringed its exclusive rights in the ’366, ’918, and ’037 Patents. That action was transferred to this court on December 16, 2003. Following Markman hearings held on September 30, 2004 and December 7, 2004, the court issued two separate orders construing the claims of the patents in suit. See November 30, 2004 Claim Construction Order (hereinafter “First Claim Construction Order”) (construing non-means-plus-function claim terms); January 7, 2005 Claim Construction Order (hereinafter “Second Claim Construction Order”) (construing means-plus-function terms).

The parties have since submitted a host of summary judgment motions in the hope of narrowing the scope of issues prior to trial. Now before the court are four of those motions, each of which arises from defendant’s contentions that the patents in suit are anticipated or rendered obvious by the prior art reference discussed above. Specifically, defendant moves for summary judgment that claims 1 and 3 of the ’366 Patent, claims 7 through 10 and 12 through 15 of the ’918 Patent, and claims 1 through 9 and 11 through 13 of the ’037 Patent are invalid as anticipated by one or more of the above-cited prior art references. Plaintiff opposes those motions and cross-moves for summary judgment of non-anticipation with respect to each of the asserted claims of the patents in suit, which include certain means-plus-function claims of the ’366 and ’918 Patents in addition to the claims at issue in defendant’s motions. In addition, plaintiff seeks summary judgment that the asserted claims of the ’366 and ’918 Patents are not invalid on the ground of obviousness and that claims 1 and 3 of the ’366 Patent satisfy the written description requirement of 35 U.S.C. § 112 ¶1. 3 The following-memorandum and order addresses the arguments raised by the parties’ motions.

LEGAL STANDARD

I. Summary Judgment

As in any other civil action, summary judgment is proper in a patent infringement action when the pleadings, discovery, and affidavits show that there is “no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(c); see also Southwall Techs., Inc. v. Cardinal IG Co., 54 F.3d 1570, 1575 (Fed.Cir.), cert. denied, 516 U.S. 987, 116 S.Ct. 515, 133 L.Ed.2d 424 (1995). Material facts are those which may affect the outcome of the case. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). A dispute as to a material fact is genuine if there is sufficient evidence for a reasonable jury to return a verdict in favor of the nonmoving party. Id.

The party moving for summary judgment bears the burden of identifying those portions of the pleadings, discovery, and affidavits that demonstrate the absence of a genuine issue of material fact. Celotex Corp. v. Catrett, 477 U.S. 317, 323, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). On an issue for which the opposing party will have the burden of proof at trial, the moving party need only point out “that there is an absence of evidence to support the non-moving party’s case.” Id. at 325, 106 S.Ct. 2548; Crown Operations Int’l, Ltd. v. Solutia, Inc., 289 F.3d 1367, 1377 (Fed.Cir.2002). On the other hand, where the moving party bears the burden of proof on an issue, it must submit evidence sufficient to establish that no reasonable jury could find against it on that issue at trial. See Frank’s Casing Crew & Rental Tools, Inc. v. Weatherford Int’l, Inc., 389 F.3d 1370, 1376 (Fed.Cir.2004); Gart v. Logitech, Inc., 254 F.3d 1334, 1339 (Fed.Cir.2001), cert. denied, 534 U.S. 1114, 122 S.Ct. 921, 151 L.Ed.2d 886 (2002).

II. Novelty and Anticipation

Section 102(a) of the Patent Act, 35 U.S.C. §. 102(a), precludes the patenting of any invention that “was known or used by others in this country, or patented or described in a printed publication in this or a foreign country” before the date of its invention. 35 U.S.C. § 102(a); Amgen, Inc. v. Hoechst Marion Roussel, Inc., 314 F.3d 1313, 1352 (Fed.Cir.2003). Similarly, section 102(b) provides that a patent claim is invalid if the patented invention is “described in a printed publication ... more than one year prior to the date of the application for patent in the United States.” 35 U.S.C. § 102(b); see Schering Corp. v. Geneva Pharm., Inc., 339 F.3d 1373, 1377 (Fed.Cir.2003), reh’g and reh’g en banc denied, 348 F.3d 992 (Fed.Cir.2003). To anticipate under either section 102(a) or section 102(b), a single prior art reference must disclose every limitation of the claimed invention. See Schering Corp., 339 F.3d at 1377 (citing Lewmar Mar., Inc. v. Barient, Inc., 827 F.2d 744, 747 (Fed.Cir.1987)). Furthermore, such disclosure must be “enabling” — i.e., it must be sufficient to permit a person having ordinary skill in the art to practice the invention. SmithKline Beecham Corp. v. Apotex Corp., 403 F.3d 1331, 1342 (Fed.Cir.2005) (quoting Minnesota Min. & Mfg. Co. v. Chemque, Inc., 303 F.3d 1294, 1301 (Fed.Cir.2002)).

Anticipation is a question of fact, id. at 1343, and the determination of whether a prior art reference is enabling “is a question of law based upon underlying factual findings,” Crown Operations Int’l, 289 F.3d at 1376. “However, without genuine factual disputes underlying the anticipation inquiry, the issue is ripe for judgment as a matter of law.” Smith-Kline Beecham, 403 F.3d at 1343. The burden of proof in all instances falls upon the party seeking to establish the invalidity of a patent claim, who “must overcome the presumption of validity in 35 U.S.C. § 282 by clear and convincing evidence.” Nystrom v. Trex Co., 374 F.3d 1105, 1117 (Fed.Cir.2004) (quoting State Contracting & Eng’g Corp. v. Condotte Am., Inc., 346 F.3d 1057, 1067 (Fed.Cir.2003)).

III. Nonobviousness

An additional prerequisite to patentability is the “nonobviousness” requirement of 35 U.S.C. § 103(a), which states:

A patent may not be obtained though the invention is not identically disclosed or described as set forth in [35 U.S.C. § 102], if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.

35 U.S.C. § 103(a). To prove that a patented invention is invalid under section 103(a), the accused infringer must identify prior art references “which alone or combined with other references would have rendered the invention obvious to one of ordinary skill in the art at the time of the invention.” AlSite Corp. v. VSI Int’l, Inc., 174 F.3d 1308, 1323 (Fed.Cir.1999) (citations omitted). In addition, if more than one prior art reference is employed, “there must be some teaching, suggestion, or motivation to combine the references.” In re Fulton, 391 F.3d 1195, 1200 (Fed.Cir.2004) (citing In re Rouffet, 149 F.3d 1350, 1355 (Fed.Cir.1998)). Although such a motivation “may flow from the nature of the problem,” there must be something in “the prior art as a whole to suggest the desirability, and thus the obviousness, of making the combination.” Ecolochem, Inc. v. Southern Cal. Edison Co., 227 F.3d 1361, 1372 (Fed.Cir.2000) (citations omitted), cert. denied, 532 U.S. 974, 121 S.Ct. 1607, 149 L.Ed.2d 473 (2001).

“Obviousness is a question of law premised on underlying findings of fact.” Eolas Techs., Inc. v. Microsoft Corp., 399 F.3d 1325, 1332 (Fed.Cir.2005) (citing Graham v. John Deere Co., 383 U.S. 1, 17-18, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966)). These underlying factual determinations include: (1) the scope and content of the prior art; (2) differences between the prior art and the claims at issue; (3) the level of ordinary skill in the art; and, if necessary, (4) secondary evidence of non-obviousness. Graham, 383 U.S. at 17-18, 86 S.Ct. 684; Para-Ordnance Mfg., Inc. v. SGS Imps. Int’l, Inc., 73 F.3d 1085, 1087-88 (Fed.Cir.1995), cert. denied, 519 U.S. 822, 117 S.Ct. 80, 136 L.Ed.2d 38 (1996). Secondary evidence of nonobviousness can include the commercial success of the invention, long-felt need, failure of others to solve the problem, licensing of the patented invention, professional recognition and approval, and copying of the invention. Graham, 383 U.S. at 17-18, 86 S.Ct. 684; Minnesota Min. & Mfg. Co. v. Johnson & Johnson Orthopaedics, Inc., 976 F.2d 1559, 1573 (Fed.Cir.1992). Like anticipation, the affirmative defense of obviousness must be established by clear and convincing evidence. See Boehringer Ingelheim Vetmedica, Inc. v. Schering-Plough Corp., 320 F.3d 1339, 1353 (Fed.Cir.2003).

IV. Written Description

In addition to the requirements that a patented invention be novel and nonobvious, the validity of a patent further depends on compliance with the written description requirements of 35 U.S.C. § 112 ¶ 1. See Space Systems/Loral, Inc. v. Lockheed Martin Corp., 405 F.3d 985, 986 (Fed.Cir.2005). Section 112 ¶ 1 states:

The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.

35 U.S.C. § 112 ¶ 1. Although a patent’s specification need not include information that is already known and available to one of ordinary skill in the art to which the patent pertains, its description of the in vention claimed must be sufficient to convey to such an ordinarily skilled artisan that the inventor was in possession of the invention on the date that the patent application was filed. Lockwood v. American Airlines, Inc., 107 F.3d 1565, 1572 (Fed. Cir.1997); Vas-Cath, Inc. v. Mahurkar, 935 F.2d 1555, 1563-64 (Fed.Cir.1991). This is achieved “by such descriptive means as words, structures, figures, diagrams, formulas, etc., that fully set forth the claimed invention.” Lockwood, 107 F.3d at 1572. The issue of compliance with the written description requirement is a question of fact. Tronzo v. Biomet, Inc., 156 F.3d 1154, 1158 (Fed.Cir.1998).

DISCUSSION

I. The ’366 Patent

A. Non-Means-Plus-Function Claims

1. Anticipation

The court first considers defendant’s assertion that the two non-means-plus-function claims of the ’366 Patent at issue in the instant action are anticipated and hence invalid. Claim 1, the sole independent claim addressed by defendant’s motion, reads as follows:

Apparatus for use with a data network and a mass storage device, comprising the combination of first and second processing units,

said first processing unit being coupled to said network and performing procedures' for satisfying requests from said network which are within a predefined non-NFS class of requests,

and said second processing unit being coupled to said network and to said mass storage device and decoding NFS requests from said network, performing procedures for satisfying said NFS requests, and encoding NFS reply messages for return transmission on said network, said second processing unit not satisfying any requests from said network which are within said predefined non-NFS class of requests.

’366 Patent, 51:35-49.

Defendant asserts that the Epoch Server described in the Lieberman and Has-kin references anticipates claims 1 as a matter of law. As noted above, the Epoch Server teaches a multiprocessor file server comprised of a front/end I/O processor running the VRTX operating system and a second processor running a standard Unix operating system, with both NFS requests and standard Unix “housekeeping chores” being handled by the second processor. Haskin at 32. Given that the Epoch Server is undisputed prior art of the ’366 Patent, the only issue raised by the parties’ motions is whether every element of the properly construed claim reads on the cited reference. See Schering Corp., 339 F.3d at 1377. Although plaintiff asserts that a number of elements of claim 1 are absent from the prior art file server, the court focuses on two of the disputed claims limitations: namely, the requirements that the claimed second processing unit “de-cod[e] NFS requests” received from a data network and “encod[e] NFS reply messages.” The court construed these terms in its November 2004 claim construction order. First Claim Construction Order at 10-12. Reading the “decoding” and “encoding” limitations in light of the (substantially identical) specifications of the ’366 and ’918 Patents, the court observed that these terms necessarily imply “a process of translation,” in which data is converted from one format to another. Id. at 11-12. Thus, the court defined “decoding NFS requests” to mean “converting NFS requests from the format received from the network to another, decoded format.” Id. at 12. Similarly, the “encoding NFS reply messages” term was construed as' “converting NFS reply messages from a decoded data format to the format used for transmission on the' network.” Id.

According to defendant, the “second processing unit” of the Epoch Server that performs the claimed “decoding” and “encoding” functions consists of the server’s front-end processor and its associated buffer memory, Ethernet interfaces, and SCSI controllers. First McKusick Decl. ¶ 43. However, the front-end processor disclosed in the Epoch Server reference performs only input-output operations, which would not on their face ordinarily be characterized as “decoding NFS requests” or “encoding NFS reply messages.” Nevertheless, defendant’s expert, Dr. Marshall McKusick, claims that the Epoch Server’s front-end processor must disclose the claimed “decoding” and “encoding” functions because the server could not function unless the front-end processor converts incoming serially “format[ted]” data packets into a parallel data “format.” First McKu-sick Decl. ¶ 48; McKusick Decl. in Support of Def.’s Reply ¶ 2 (hereinafter “Third McKusick Decl.”). According to Dr. McKusick, this process of serial-to-parallel conversion amounts to “decoding NFS requests” and “encoding NFS reply messages.” Id.

The court assumes for the purpose of deciding the instant motions that such a serial-to-parallel conversion process is inherently disclosed in the Epoch Server reference. However, even granting defendant the benefit of this assumption, a jury would be compelled to find that the Epoch Server does not teach the claimed decoding and encoding functions. Defendant’s gratuitous use of the word “format” notwithstanding, converting serially transmitted data into a parallel data stream simply does not amount to “decoding” NFS requests or “encoding” NFS reply messages. As the court has construed these terms, they are best analogized to the process of translating a word from one language to another. By way of contrast, the serial-to-parallel conversion performed by the Epoch Server’s front-end processor is more closely analogized to reorganizing or rearranging a language’s constituent components, those components being words in the case of a spoken language and data packets in the context of the patented invention. Such a process is fundamentally different from the procedures that are described in the ’366 Patent’s specification, which involve accepting an incoming NFS packet, reading its protocol layers, and converting the NFS-formatted packet to the LNFS format that the preferred embodiment of the patented file server uses for internal communications among its constituent processors. It is also inconsistent with the process of translation that the court has construed to be included in the “decoding” and “encoding” functions recited in claim 1. Thus, seeing that defendant has identified no other feature of the prior art file server that might teach the claimed “decoding” and “encoding” functions, the court holds as a matter of law that the Epoch Server’s front-end processor does not disclose a “second processing unit” meeting all of the limitations of claim l. 4

While this conclusion compels the court to deny defendant’s motion for summary judgment, defendant’s memorandum in opposition to plaintiffs motion also identifies the Coulouris textbook as anticipatory prior art that precludes entry of summary judgment in plaintiffs favor. Admittedly, Coulouris teaches a number of distributed file system architectures that describe the partitioning of different operating system functions among multiple processing units. See generally Coulouris at 25-32, 107; McKusick Decl. in Support of Def.’s Opp’n ¶¶ 22-23 (hereinafter “Second McKusick Deck”). The textbook also provides a general description of the NFS operating system. See Coulouris at 264-66. These, however, are distinct teachings in a lengthy introductory text on distributed computing systems. While both of these disclosures may of course be weighed in considering whether the prior art renders the patent’s claims obvious, it is generally inappropriate to combine such distinct teachings for the purpose of determining the novelty of a patented invention, even though they may be encompassed within a single prior art reference. See Sandisk Corp. v. Lexar Media, Inc., 91 F.Supp.2d 1327, 1336 (N.D.Cal.2000) (Breyer, J.) (observing that there can be no anticipation of a patent “[ujnless all the elements are found in a single piece of prior art in exactly the same situation and united the same way to perform the identical function”); see also Brown v. 3M, 265 F.3d 1349, 1351 (Fed.Cir.2001) (“To anticipate, every element and limitation of the claimed invention must be found in a single prior art reference, arranged as in the claim.”), cert. denied, 535 U.S. 970, 122 S.Ct. 1436, 152 L.Ed.2d 380 (2002). Although it may be true that the Coulouris reference teaches both a distributed file server and the NFS filesystem, there is nothing in its description of the latter to suggest the segregation of non-NFS and NFS functionality between multiple processing units that is required by claim 1. Indeed, if anything, the description of NFS in Coulouris teaches away from this aspect of the patented invention, noting that “[a]ny computer running NFS software can become a server machine by making some of its file system available for remote clients to mount.” Coulouris at 265. This suggests the duplication of file system functionality on each of the processing units rather than the segregation of that function onto a separate processing unit that does “not satisfy! ] any requests from [a data] network which are within [a] predefined non-NFS class of requests,” as would be required to meet every limitation of the patented invention. Thus, the court concludes as a matter of law that Coulour-is does not teach the file server defined by claim 1.

In light of this conclusion, plaintiff is entitled to summary judgment that claim 1 is not anticipated by the cited prior art references. Moreover, because claim 3 depends on claim 1, it necessarily shares each of the aforementioned claim limitations. The court therefore holds as a matter of law that neither the Epoch Server nor the Coulouris reference anticipates claims 1 or 3 of the ’366 Patent.

2. Obviousness

Having held that the asserted non-means-plus-function claims of the ’366 Patent are not anticipated, the court must address plaintiffs contention that it is entitled to summary judgment of nonobviousness with respect to those claims. As previously noted, the question of a patent claim’s obviousness turns on whether an accused infringer can show by clear and convincing evidence that one or more prior art references, either alone or in combination, would have rendered the invention obvious to one of ordinary skill in the art on the patent’s critical date. See Al-Site, 174 F.3d at 1323. Furthermore, to the extent that an obviousness defense relies upon multiple prior art references, the defendant bears the burden of establishing that an ordinarily skilled artisan would have been motivated to combine those references prior to the date of invention. See Ecolochem, 227 F.3d at 1372.

The court first considers whether claim 1 could be rendered obvious by some combination of the prior art references upon which defendant relies, namely, the Sandberg, Palowski, and V-System references in addition to the Epoch Server and Coulouris textbook. Taken as a whole, these references make clear that the NFS file system was well-known in the prior art extant on the ’366 Patent’s critical date. See generally Sandberg at 10399-411; Coulouris at 264-66. The prior art also teaches a number of multiprocessor file server architectures, at least some of which contemplate the patented invention’s separation of file system functions and non-file system functions onto different processing units. See V-System Manual at 1-3 to 1-4; Coulouris at 25-32. Setting aside for the moment the question of whether there is substantial evidence of a motivation to combine these references, the teachings set forth in the cited prior art would appear to give rise to a genuine issue of material fact with respect to the obviousness of claim 1.

Plaintiff nevertheless argues that summary judgment of nonobviousness is appropriate because none of these references, singularly or in combination, suggest a multiprocessor architecture for handling NFS requests from a data network in which a “predefined non-NFS class of requests” is processed by one of the processing units but “not satisfied” by the other unit. This is an unduly narrow view of the prior art. One example of such a file server architecture appears in Coulouris’ discussion of “hybrid” distributed networks. Coulouris at 26, 29-30 & fig. 2.4. In such a hybrid network, file system requests from client computers on the network are directed to one or more file servers, while other “pool processor” computers handle other requests from client workstations (for example, executing applications that are too resource-intensive to be executed on client computers). Id. Yet another example is the V-System reference, which teaches that file storage-related tasks may be processed by a workstation server operating on a distributed network. V-System Manual at 1-3 to 1-4. Both of these reference provide compelling evidence that the segregation of file system and non-file system tasks onto separate servers would have been known to an ordinarily skilled artisan on the patent’s critical date.

Plaintiff also argues that the claimed “decoding” and “encoding” functions are not taught by any of the cited prior art references. However, these terms, while narrower than the almost absurdly broad “performing procedures” element of claim 1, are nonetheless expansive in scope. Indeed, given that the NFS filesystem employs a data encoding standard known as “external data representation” (“XDR”) for transferring data between different computer architectures, ’366 Patent, 6:29— 52, the claimed “decoding” and “encoding” functions could potentially read a wide range of processing units that respond to NFS requests from a data network. Thus, it can be hardly be said as a matter of law that such limitations are absent from the prior art.

Finally, despite plaintiffs assertion to the contrary, the court cannot rule out the possibility that a reasonable jury would find that an ordinarily skilled artisan would have had a motivation to combine the above-cited references in order to achieve the benefits realized by the patented invention. The prior art references that defendant identifies all pertain to the same art as the invention, and like the ’366 Patent, the references are generally directed at the problem of improving the performance of file servers operating in a Unix-based networked computing environment. Moreover, the particular means that the ’366 Patent’s inventors choose to achieve such performance improvements, the partitioning of operating system functions onto multiple processors, was also known in the prior art. For example, in discussing the advantages of the IBM mainframe port of NFS taught in Palowski, the authors observe that the offloading of “user interface” and “data manipulation” functions onto a Sun front-end processor frees the mainframe computer’s resources to perform file system-related “data management functions.” Palowski at 291. Similarly, the V-System Manual teaches the value of partitioning operating system functions onto separate workstation “servers.” V-System Manual at 31-8. In light of such evidence, it is impossible to conclude as a matter of law that an ordinarily skilled artisan would have lacked the motivation to combine the cited references in a manner that would have rendered the patented invention obvious. The court therefore denies plaintiffs motion for summary judgment with respect to claim 1.

The other non-means-plus-function claim of the ’366 Patent at issue is claim 3. Claim 3 narrows the scope of claim 1 by requiring the claimed first processing unit to include a “general purpose operating system,” while prohibiting the second processing unit from having such a general purpose operating system. ’366 Patent, 51:53-56. Such a configuration is taught by the Epoch Server, which discloses a multiprocessor architecture comprised of one processor running a “low-overhead” VRTX processor and a second processor running the standard Unix operating system. Lieberman at 37; Haskin at 32; see also Third McKusick Decl. ¶ 5 (contrasting VRTX with a general purpose operating system such as Unix). Furthermore, the V-System Manual teaches the advantages that can be achieved by removing the unnecessary features of a complete operating system from the individual processing units of a multiprocessor file server. V-System Manual at 1-3, 31-8. Such known advantages of “low-overhead” file server architectures could be found to provide a motive to combine a special-purpose operating system with the above-cited prior art references. Accordingly, plaintiffs motion for summary judgment of nonobviousness must also be denied with respect to claim 3.

3. Written Description

The next issue raised by the parties’ motions is whether claims 1 and 3 of the ’366 Patent satisfy the written description requirement of 35 U.S.C. § 112 ¶ 1. Plaintiff moves for summary judgment on this issue, asserting that defendant has not established a triable issue of fact as to whether the inventors of the ’366 Patent adequately described the claimed invention in their patent application. Defendant opposes this motion.

The requirement for establishing invalidity under section 112 ¶ 1 is straightforward: a claim will be invalid for lack of an adequate written description if the inventor fails to describe every limitation of the claimed invention in the patent’s specification. See Lockwood, 107 F.3d at 1572; see also PIN/NIP, Inc. v. Platte Chem. Co., 304 F.3d 1235, 1247-48 (Fed.Cir.2002). Defendant’s theory of invalidity is premised upon the claim limitations requiring that the patented file server be comprised of first and second processing units, with both of those processing units being “coupled to [a data] network.” Defendant argues the ’366 Patent’s specification fails to describe a file server having these limitations because the sole embodiment disclosed by the inventors describes an operating system in which only one of the processing units (the file server) is directly connected to a data network. ’366 Patent, 7:64-8:18 & fig. 2. The other processing unit (the host processor) accesses the network indirectly via the file server and a VME bus. Id. Thus, according to defendant, claims 1 and 3 are invalid under section 112 ¶ 1.

Implicit in this argument is the assumption that claim l’s “coupled to” term must be construed to mean “directly coupled to” or “directly connected.” However, there is no basis for reading such a limitation into the claim. As the District of Delaware has observed in construing the claims of a patent for a computer graphics chip, “the ordinary and accustomed meaning of the term ‘couple,’ even when used in an electronics context[,] does not solely mean ‘directly coupled.’ ” Silicon Graphics, Inc. v. n Vidia Corp., 58 F.Supp.2d 331, 346 (D.Del.1999). Defendant offers no reason why this ordinary meaning should not apply to the “coupled to” term of the ’366 Patent. Indeed, defendant’s theory of invalidity demonstrates that a narrower “directly connected” construction of the “coupled to” claim element would fail to read on the invention’s preferred embodiment. Such a result is “rarely, if ever, correct and would require highly persuasive evidentiary support.” Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1583 (Fed.Cir.1996). No such support has been offered by defendant here. Thus, properly construed, the “coupled to” limitation of claim 1 (and hence of dependent claim 3) is adequately described in the ’366 Patent’s specification. Accordingly, the court grants plaintiffs motion for summary judgment that those claims satisfy the written description requirement.

B. Means-Plus-Function Claims

Although the preceding discussion resolves the issues raised by defendant’s motion for summary judgment of invalidity with respect to the claims of the ’366 Patent, plaintiffs motion also raises issues of anticipation and obviousness concerning the patent’s asserted means-plus-function claims (specifically, claims 5 through 9). The court considers those issues below.

Plaintiff first argues that it is entitled to summary judgment that the asserted means-plus-function claims of the ’366 Patent are not anticipated by the Coulouris or Epoch Server references, each of which defendant cites as anticipatory prior art. Under 35 U.S.C. § 112 ¶ 6, any element of a patent’s claims that is drafted in “means-plus-function” format must be “construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.” 35 U.S.C. § 112 ¶ 6. The anticipation of a means-plus-function claim element is governed by the same standard that applies in determining whether such an element is literally infringed. See Transclean Corp. v. Bridgewood Servs., Inc., 290 F.3d 1364, 1372 (Fed.Cir.2002). Consequently, to support a finding of anticipation, there must be sufficient evidence to establish that “the assertedly equivalent structure performs the claimed function in substantially the same way to achieve substantially the same result as the corresponding structure described in the specification.” Frank’s Casing Crew & Rental Tools, 389 F.3d at 1378 (quoting Odetics, Inc. v. Storage Tech. Corp., 185 F.3d 1259, 1267 (Fed.Cir.1999)).

Applying this standard to the ’366 Patent’s means-plus-function claims, the court begins its analysis with claim 5, which claims the following:

A network file server for use with a data network and a mass storage device, said network file server including a first unit comprising:

means for decoding NFS requests from said network;

means for performing procedures for satisfying said NFS requests, including accessing said mass storage device if required; and

means for encoding any NFS reply messages for return transmission on said network,

said first unit lacking means in said first unit for executing any programs which make calls to any general purpose operating system.

’366 Patent, 52:44-55. As the court has already held that the claimed “decoding” and “encoding” functions are not taught by the Epoch Server’s front-end processor, claim 5, like claims 1 and 3 of the ’366 Patent, cannot be anticipated by that reference. See Transclean, 290 F.3d at 1372 (citation omitted) (“To anticipate a claim reciting a means-plus-function limitation, the anticipatory reference must disclose the recited function identically.”). The court thus need only consider whether the Coulouris textbook discloses every functional and structural limitation of the invention.

As an initial matter, it should be noted that the court’s discussion of claims 1 and 3 makes clear that Coulouris teaches the claimed decoding, encoding, and performing procedures functions. However, to prove that claim 5 is anticipated, defendant must also show that these functions' — in particular, the claimed “decoding” and “encoding” functions — are carried out by means equivalent to those disclosed in the specification. See 35 U.S.C. § 112 ¶ 6. The first step of this inquiry is to identify the means for carrying out these functions that are disclosed in the patent’s specification, a task that the court performed in its January 2005 claim construction order. In that order, the court, relying on WMS Gaming, Inc. v. International Game Technology, Inc., 184 F.3d 1339 (Fed.Cir.1999), held that the claimed “decoding” and “encoding” functions correspond to a microprocessor that is programmed to carry out the procedures performed by the “network controller” disclosed in the patent’s specification and the equivalents of that structure. Second Claim Construction Order at 9-11. With respect to the claimed “decoding” function, these procedures include processing all of the necessary layers of the NFS protocol stack and converting the fully processed NFS request into the LNFS format. ’366 Patent, 9:34^5. Similarly, the structure corresponding to the “encoding” function is a microprocessor programmed to reverse those steps. Id. at 9:46-47. Consequently, to teach claimed decoding and encoding means, Coulouris must disclose either a processor that is programmed to process the full NFS stack, to convert fully processed NFS requests into LNFS format, and to reverse those processes in order to generate reply messages, or, alternatively, a processor that is programmed to execute an equivalent algorithm.

The court assumes without deciding that LNFS is equivalent to any format used for internal communications among the components of a multiprocessor file server and is thus disclosed in the cited prior art references. 5 However, even granting de fendant the benefit of this assumption, there is nothing in Coulouris to suggest that the claimed “decoding” and “encoding” functions are performed by a server lacking the means to execute a general purpose operating system. While defendant cites processors that execute certain lower-level protocols such as the transport or presentation layers as evidence that Coulouris teaches every limitation of claim 5, see Second McKusick Deck ¶¶ 20,' 28, these citations are insufficient as a matter of law to anticipate the “means for decoding” and “means for encoding” elements because the disclosed prior art processors do not process the full NFS protocol stack. Moreover, to the extent that Coulouris’ discussion of the NFS filesystem does disclose a file server that is capable of processing all NFS protocol layers, see generally Coulouris at 264-69, there is nothing in that discussion to suggest, either explicitly or inherently, the partitioning of operating system components that is required to meet claim 5’s “lacking means ... for executing any programs which make calls to any general purpose operating system” limitation. Thus, seeing that defendant has not identified any algorithm taught by Coulouris that might be equivalent to these procedures, the court is compelled to conclude that the Coulouris reference does not disclose a processor lacking a general purpose operating system that includes the claimed means for decoding NFS requests and encoding NFS reply messages.

For this reason, Coulouris, like the Epoch Server reference, could not reasonably be found to teach every element of claim 5 of the ’366 Patent. Furthermore, because the remaining means-plus-function claims at issue, claims 6 though 9, are dependent claims of claim 5, they are likewise not anticipated by the Epoch Server or Coulouris references. The court therefore holds as a matter of law that the asserted means-plus-function claims of the ’366 Patent are not anticipated.

The final issue raised by plaintiffs motion with respect to the ’366 Patent is whether the patent’s means-plus-function claims are nonobvious as a matter of law. Here too, the court focuses on claim 5, with the dispositive issue again being the requirement that the prior art teach a microprocessor file server programmed to perform the claimed “decoding” and “encoding” functions by the same means that are disclosed in the patent’s specification, while at the same time lacking a general purpose operating system. Defendant has presented substantial evidence that each of these elements is taught by the prior art. Indeed, the procedures for processing NFS filesystem requests were known to ordinarily skilled artisans on the patent’s critical date, and the most elementary instantiations of NFS teach a processor that executes all of the protocols necessary to process that request. See generally Sand-berg at 10400-02. Furthermore, the Pa-lowski and Epoch Server references respectively teach the offloading of non-NFS functionality onto one of the processing units of a multiprocessor file server and a file server that includes a processing unit running'a “low-overhead” (i.e., simplified) operating system for performing I/O-related functions. See Palowski at 291; Haskin at 32.

That being the case, what is completely absent from the prior art references is any suggestion to combine such teachings in a manner that would render the patented invention obvious. For example, in Epoch Server, the “low-overhead” VRTX operating system disclosed in that reference runs on the front-end processor rather than the Unix file processor. Haskin at 32. It is the latter processing unit (which does have a complete operating system) that processes the NFS protocol stack and hence performs the “decoding” and “encoding” functions recited in claim 5. Id. Similarly, the V-System Manual, which defendant also cites as evidence that would support a finding of obviousness, does not teach anything resembling the claimed means for decoding and encoding. Indeed, it discloses a processor for handling NFS requests that runs a complete Unix operating system, see V-System Manual at 43-1 to 43-4, a disclosure which, if anything, teaches away from the patented invention. Thus, given the absence of any evidence suggesting that a person of ordinary skill in the art would have had a motive to combine the above-cited references on the patent’s critical date, the court holds that no reasonable jury could find claim 5 of the ’366 Patent to be obvious. 6 The court therefore grants plaintiffs motion for summary judgment of nonobvi-ousness (and hence validity) with respect to claim 5 and its dependent claims.

II. The ’918 Patent

A. Norir-Means-Plus-Function Claims

The court next turns to defendant’s argument that the asserted non-means-plus-function claims of the ’918 Patent are anticipated. The first of the two independent claims at issue, claim 7, reads as follows:

A network file server for use with a data network and a mass storage device, said network file server comprising:

a network control module, including a network interface coupled to receive file system requests from said network;

a file system control module, including a mass storage device interface coupled to said mass storage device; and

a communication path coupled directly between said network control module and said file system control module, said communication path carrying file retrieval requests prepared by said network control module in response to received file system requests to retrieve specified retrieval data from said mass storage device,

said file system control module retrieving said specified retrieval data from said mass storage device in response to said file retrieval requests and returning said specified retrieval data to said network control module,

and said network control module preparing reply messages containing said specified retrieval data from said file system control module for return transmission on said network.

’918 Patent, 127:40-62.

The first prior art reference identified in defendant’s motion is Sandberg’s 1986 article discussing the development and implementation of the Sun NFS filesystem. While the Sandberg reference focuses on the development of the NFS protocols, it also describes examples of its implementa tion in a number of computing environments, including a System V.2 file server comprised of a VAX 750 processor and an “Excelan” networking board. Sandberg at 10410. Defendant asserts that this “port” of the NFS filesystem to a System V.2 server meets every limitation of claim 7, with the Excelan board corresponding to the claimed network control module and the VAX 750 embodying the file system control module.

The court begins its analysis with the network control module. Based on the court’s construction of the “module” claim term and the various functions recited in claim 7, the patented network control module reads on any “distinct hardware unit” performing network control functions — i.e.; a distinct hardware unit that prepares file retrieval requests in response to file system requests from client computers and prepares reply messages for return transmission of data to those clients. See First Claim Construction Order at 18. There is no dispute that the Excelan board component of the prior art server is a distinct hardware unit. As for the functions that the claimed network control module performs, the court’s claim construction order expressly rejected a construction of the “file retrieval requests prepared” and “preparing reply messages” terms that would require the network control module to process all layers of the NFS protocol stack, thereby making clear that those claim limitations read on a processor that executes only lower-level NFS protocols such as the network and transport layers. Id. at 13-15. In the Sandberg reference, the Excelan board processes such lower-level layers of the NFS protocol stack. See Sandberg at 10410 (stating the Exce-lan board “handles Ethernet, IP, and UDP requests”). Accordingly, Sandberg clearly teaches the functional as well as the structural limitations of the claimed network control module.

With respect to the file system control module, the analysis is more difficult in that a number of features of the patented invention are either vaguely described or completely absent in Sandberg’s express teachings. Plaintiff argues that this compels the court to conclude that Sandberg does not anticipate claim 7. Federal Circuit law is to the contrary, however. It is true that “invalidity by anticipation requires that the four corners of a single, prior art document describe every element of the claimed invention.” Advanced Display Sys., Inc. v. Kent State Univ., 212 F.3d 1272, 1282 (Fed.Cir.2000) (citations omitted), cert. denied, 532 U.S. 904, 121 S.Ct. 1226, 149 L.Ed.2d 136 (2001). Nevertheless, this inquiry must be undertaken from the perspective of a person having ordinary skill in the pertinent art. See id. (observing that a prior art document must “describe every element of the claimed invention, either expressly or inherently, such that a person of ordinary skill in the art could practice the invention without undue experimentation”); Glaxo, Inc. v. Novopharm Ltd., 52 F.3d 1043, 1047 (Fed.Cir.) (citation omitted) (observing that the disclosure in the prior art “need not be express, but may anticipate by inherency where it would be appreciated by one of ordinary skill in the art”), cert. denied, 516 U.S. 988, 116 S.Ct. 516, 133 L.Ed.2d 424 (1995). The tension between the so-called “four comers” rule for finding anticipation and the role of the ordinarily skilled artisan in that analysis was aptly described by the Federal Circuit in Elan Pharmaceuticals, Inc. v. Mayo Foundation for Medical Education & Research, 346 F.3d 1051 (Fed.Cir.2003) (en banc). Discussing the proper standard to apply in determining the content of a prior art reference, the Elan court observed:

It is well settled that prior art under 35 U.S.C. § 102(b) must sufficiently describe the claimed invention to have placed the public in possession of it. Such possession is effected if one of ordinary skill in the art could have combined the publication’s description of the invention with his own knowledge to make the claimed invention. Accordingly, even if the claimed invention is disclosed in a printed publication, that disclosure will not suffice as prior art if it is not enabling. It is not, however, necessary that an invention disclosed in a publication shall have actually been made in order to satisfy the enablement requirement.

Id. at 1055 (quoting In re Donohue, 766 F.2d 531, 533 (Fed.Cir.1985)). Thus, the question here is not whether the Sandberg reference expressly discloses every element of the file system control module of the claimed invention, but rather whether its description of the System V.2 port of NFS would have been sufficient to permit an ordinarily skilled artisan to practice the patented file server without undue experimentation. See Advanced Display Sys., 212 F.3d at 1282.

Of course, simply stating this rule does not answer the question of what an ordinarily skilled artisan would have understood to be inherently present in the Sandberg reference as of the ’918 Patent’s critical date. To answer this question, the Federal Circuit has repeatedly recognized that it is appropriate for a district court to rely on extrinsic evidence, including expert testimony. See, e.g., Telemac Cellular Corp. v. Topp Telecom, Inc., 247 F.3d 1316, 1328 (Fed.Cir.2001). The parties both submit declarations of experts whose qualifications in the art of file server design are undisputed, with plaintiff and defendant respectively relying on the opinions of Dr. Phil Faillace and Dr. Marshall McKusick. While it is hardly surprising that the parties’ expert witnesses reach different conclusions as to whether the Sandberg reference anticipates the ’918 Patent’s claims, what is significant about the evidence that they present is the different methods that Dr. Faillace and Dr. McKusick employ in reaching their respective conclusions. Indeed, it is telling that Dr. Faillaee’s declarations focus almost exclusively on the absence of express disclosure of the claim elements in the Sandberg reference, carefully eschewing any mention of whether an ordinarily skilled artisan would recognize any particular claim limitation as being disclosed by its teachings. See generally Faillace Decl. in Support of Pl.’s Mem. P. & A., Exh. B at B-4-1 to B-4-3 (hereinafter “First Faillace Decl.”); Faillace Deck in Support of Pl.’s Opp’n ¶¶ 25-31 (hereinafter “Second Faillace Deck”). Instead, Dr. Faillace merely recites the applicable legal standard, concluding that numerous features of the claimed invention are neither expressly nor inherently present in the Sandberg reference. See id. Under Federal Circuit law, a reasonable jury could not be persuaded by such “evidence” of a claim’s validity. See, e.g., Dynacore Holdings Corp. v. U.S. Philips Corp., 363 F.3d 1263, 1278 (Fed.Cir.2004) (citations omitted) (noting, in the context of infringement, that “an expert’s unsupported conclusion on the ultimate issue ... is insufficient to raise a genuine issue of material fact”). Consequently, Dr. Faillace’s unsubstantiated legal conclusions cannot by themselves defeat defendant’s motion for summary judgment of anticipation.

In contrast, Dr. McKusick describes with reasonable particularity how the various elements of the patented invention are taught by the Sandberg reference, thereby providing evidence of how a person of ordinary skill in the art would have viewed the reference’s admittedly brief description of the VAX 750 port of NFS in the late 1980s. See generally First McKusick Deck ¶¶ 36-41. It is true that such expert testimony, if unsupported by documentary evidence, “cannot supplant the requirement of anticipatory disclosure in the prior art reference itself.” Motorola, Inc. v. Interdigital Tech. Corp., 121 F.3d 1461, 1473 (Fed.Cir.1997) (citing Jamesbury Corp. v. Litton Indus. Prods., Inc., 756 F.2d 1556, 1563 (Fed.Cir.1985)). Here, however, defendant has supplemented Dr. McKusick’s declarations with hardware documentation describing the components of the prior art server, see Third McKusick Dec!., Exhs. B-D. Taken together with Dr. McKusick’s expert opinion, these documents constitute competent evidence that is relevant in determining whether the features of the patented invention would have been understood to be inherently disclosed by Sandberg.

With these considerations in mind, the court turns to the task of comparing the elements of claim 7’s “file system control module” with the VAX 750 component of the prior art file server disclosed in the Sandberg reference. Again, the court begins this inquiry with the construction of the claim limitation at issue. As noted above, the court has construed the “module” claim element to read on any “distinct hardware unit.” First Claim Construction Order at 18. However, the court has yet to address the proper construction of the “file system control module” term itself, which plaintiff now seeks to limit to a “distinct hardware unit dedicated principally to the control of a file system.” First Faillace Deck, Exh. B at B-5-1. In support of this construction, plaintiff cites numerous passages from the ’918 Patent’s specification, each of which describe the invention’s preferred embodiment. See Second Faillace Deck ¶¶ 13-18. This evidence simply does not permit the court to read a “dedicated principally” limitation into the “file system control module” term. Indeed, having routinely insisted upon construing the terms of the patents in suit according to their ordinary meaning during claim construction proceedings, plaintiff hardly needs to be reminded that “the presumption of ordinary meaning cannot be rebutted ‘simply by pointing to the preferred embodiment or other structures or steps disclosed in the specification or prosecution history.’” Playtex Prods., Inc. v. Procter & Gamble Co., 400 F.3d 901, 908 (Fed.Cir.2005) (quoting CCS Fitness, Inc. v. Brunswick Corp., 288 F.3d 1359, 1366 (Fed.Cir.2002)). That is precisely what plaintiff has attempted to do here. Accordingly, so long as the VAX 750 processor disclosed in Sandberg is a distinct hardware unit that performs file system functions — and there is no dispute that it is — it is capable of satisfying the “file system control module” limitation of the patented invention.

That being the case, claim 7 imposes a number of structural and functional limitations on the claimed file system control module, each of which must be present in the VAX 750 processing unit of the prior art file server if anticipation is to be established. The court first considers the structural elements of the claim, beginning with a “mass storage device interface coupled to [a] mass storage device.” Plaintiff argues that this limitation is absent, correctly observing that Sandberg makes no express mention of a mass storage device or mass storage device interface in its description of the VAX 750. However, in taking this position, plaintiff ignores undisputed evidence that it would have been impossible for the disclosed System V.2 port of NSF to perform “file system functions” such as creating or removing directories if it were not able to access a mass storage device. See Third McKusick Deck ¶ 11. Such functions are explicitly taught by the Sandberg reference, see Sandberg at 10403, and would in any event have been known to persons having ordinary skill in the art on the critical date of the ’918 Patent, see ’918 Patent, 5:30-7:13 (describing prior art file servers). See also Third MeKusick Deck ¶ 11 (discussing prior art knowledge of NFS). As for the particular mass storage device disclosed, Dr. McKu-sick identifies the device as a hard disk and attests that one of ordinary skill in the art would have recognized it as such in 1989. See First MeKusick Decl. ¶ 38. Moreover, he supports this opinion with documentary evidence that the VAX 750 typically includes one to two tape drives and two to six disks employing a Unibus or Massbus interface. Third MeKusick Deck, Exh. D. In contrast, the only “evidence” that plaintiff submits to rebut Dr. McKu-sick’s declarations consists of the unsubstantiated legal conclusions of Dr. Faillace. For the reasons discussed above, such evidence, if it can be called that, is insufficient to rebut defendant’s clear and convincing evidence of the contents of the prior art on the patent’s critical date. Thus, a reasonable jury would be compelled to conclude that the “mass storage device” and “mass storage device interface” elements of the claimed invention are present in the VAX 750 component of the prior art file server.

The next element of claim 7 is “a communication path coupled directly between [the] network control module and [the] file system control module.” Here, one hardly needs to resort to expert testimony to conclude that there must be some path by which client requests from the network are transmitted from the Excelan board to VAX 750 for processing of the higher-level protocols in the NFS stack. MeKusick identifies this communication path as a 32-bit Unibus and further notes that such a bus provides a direct connection between the Excelan board and the VAX 750. First MeKusick Deck ¶¶ 38-39; see also Third MeKusick Deck, Exh. C at 1-7 (describing the bus interfaces of the Excelan board); Third MeKusick Deck, Exh. D (describing disk or tape drives connected to a VAX 750 via a Unibus or Massbus interface). Thus, uncontroverted evidence establishes that the “communication path” element of claim 7 also reads on the System V.2 port of NFS taught by Sandberg.

In short, each of the structural limitations of the claimed file system control module are present in the Sandberg reference. That leaves the court to consider whether Sandberg teaches the functions performed by the file system control module, turning first to the preparing file retrieval requests (i.e., “file retrieval requests prepared”) and “preparing reply messages” limitations. Here too, the evidence that these claim elements are present in the prior art file server is uncontro-verted. Like any instantiation of NFS, the System V.2 port performs the function of processing NFS protocols in response to a request from a client computer on a data network. First MeKusick Deck ¶¶ 39^40; Sandberg at 10400-02. Among such functions are reading and writing data from the server. Id. Indeed, Sandberg expressly identifies “commit[ting] ... modified data to stable storage” (i.e., writing data to a mass storage device) as one of the procedures that NFS servers must perform in order to service an NFS request. Sand-berg at 10402. This is but one example of the numerous protocols disclosed in the Sandberg reference that fall within the broad scope of the “file retrieval requests prepared” and “preparing reply messages” terms. See id. at 10403-06 (describing various NFS protocols). For similar reasons, a person of ordinary skill in the art would have understood Sandberg’s disclosure of the NFS protocol procedure “read” to teach a file server that “retriev[es] specified retrieval data from [a] mass storage device.” Id. at 10401; Third MeKusick Deck ¶ 11. Finally, in order to perform the functions disclosed in Sandberg, the retrieved data would necessarily be returned to client computers on the network via the Excelan board. First MeKusick Deck ¶ 41. This teaches a file system control module “returning ... specified re trieval data to [a] network control module.” The court therefore holds that any reasonable jury would conclude that every functional limitation of claimed file system control module is present in the VAX 750 component of the prior art server.

Having addressed each of the claim limitations at issue, the court holds that there is no genuine dispute of material fact as to whether the System V.2 port of NFS taught by Sandberg anticipates claim 7. In light of this conclusion, the anticipation of the other asserted non-means-plus-function claims of the ’918 Patent can be dealt with summarily. With respect to claims 8 through 10, each of which depends on claim 7, Dr. Faillace’s declarations add nothing to the legally insufficient “evidence” that plaintiff has submitted regarding the validity of claim 7. See First Fail-lace Decl., Exh. B at B-4-3 to B-4-4; Second Faillace Deck ¶¶ 29-30. Accordingly, this court finds that defendant has established by clear and convincing evidence that claims 7 through 10 are anticipated by Sandberg and that it is entitled to summary judgment of anticipation with respect to those claims. Furthermore, because claims 12 through 15 .are method claims that are comprised of substantially identical elements to the devices claimed by claims 7 through 10, they are similarly anticipated by the prior art file server disclosed in Sandberg. The court therefore grants defendant’s motion for summary judgment that claims 7 through 10 and claims 12 through 15 of the ’918 Patent are anticipated by the Sandberg reference and are thus invalid under 35 U.S.C. § 102(b). 7

The court next turns to the means-plus-function claims of the ’918 Patent and plaintiffs motion for summary judgment of validity with respect to those claims. Each of the asserted means-plus-function claims (claims 1, 3, 5 and 9) includes the same “means for decoding,” “means for encoding,” and “lacking means ... for executing any programs which make- calls to any general purpose operating system” limitations found in claim 5 of the ’366 Patent. Thus, having held that the ’366 Patent’s mean-plus-function claims are neither anticipated nor rendered obvious by the prior art, the court must reach the same conclusion with the means-plus-function claims of the ’918 Patent. The court therefore grants plaintiffs motion for summary judgment of validity with respect to those claims.

III. The ’037 Patent

The final patent in suit, the ’037 Patent, claims a computer system for implementing multiprocessor operating system architectures such as those claimed by the ’366 and ’918 Patents. The sole issue raised by the parties’ motions for summary judgment with respect to the ’037 Patent is whether claims 1 through 9 and claims 11 through 13 are anticipated. The court first considers claim 1, which recites:

A computer system employing'a multiple facility operating system architecture, said computer system comprising:

a) a plurality of processor units provided to co-operatively execute a predetermined set of operating system peer-level facilities, wherein each said processor units is associated with a respective different one of said operating system peer-level facilities and not another of said operating system peer level facilities, and wherein each of said operating system peer-level facilities constitutes a respective separately executed software entity which includes a respective distinct set of peer-level facility related functions, each said processor unit including:

i) a processor capable of executing a control program; and

ii) a memory store capable of storing said control program, said processor being coupled to said memory store to obtain access to said control program,

said memory store providing for the storage of a first control program portion that includes a one of said respective distinct sets of operating system peer-level facility related functions and that corresponds to a one of said predetermined operating system peer-level facilities, and a second control program portion that provides for the implementation of a multi-tasking interface function, said multi-tasking interface function being responsive to control messages for selecting for execution a one of said peer-level facility related functions of said one of said predetermined operating system peer-level facilities and responsive to said one of said predetermined operating system peer-level facilities for providing control messages to request or in response to the performance of said predetermined peer-level facility related functions of another operating system peer-level facility; and

b) a communications bus that provides for the interconnection of said plurality of processor units, said communications bus transferring said control messages between the multi-tasking interface functions of said predetermined set of operating system peer-level facilities.

’037 Patent, 55.T8-58.

In moving for summary judgment of invalidity, defendant asserts that each of the elements of claim 1 is present in the V-System Manual. As previously noted, this reference describes a “distributed” file server in which various functions of the operating system are performed by separate workstations, each of which controls a separate resource on the network. See generally V-System Manual at 1-2 to 1-3. The operating system running on each of these workstations is divided into two major components, the first component being a “host kernel” that “provides process management, interprocess communications, and low-level device management facilities,” and the second being a workstation-specific component that implements other operating system services outside the kernel. Id. at 1-3. Communication between each of the host kernels, which are collectively referred to as the “distributed kernel,” is provided via an “inter-kernel protocol,” which the V-System Manual describes as “a reliable request-response protocol, intermediate in complexity between conventional datagram and virtual circuit protocols.” Id.

Under defendant’s theory of invalidity, each of the workstations that comprise the distributed V-System server is a “processing unit” capable of executing one of the “operating system peer-level facilities” recited in claim 1. In addition, within each of these “peer-level facilities,” defendant further identifies the host kernel portion of the workstation’s operating system as corresponding to the “multi-tasking interface function” claim element and the workstation-specific pi'oeesses executed outside the kernel as being identical to the claimed “control program.” While plaintiff objects to this theory on various and sundry grounds, the court need only consider the “multi-tasking interface function” term for purposes of the instant motions.

In its November 2004 claim construction order, the court construed this term as a matter of law, defining the claimed “multitasking interface function” as “software that is tailored to each operating system peer-level facility that supports direct communication with other peer-level facilities and capable of managing at least two concurrent or interleaved tasks.” First Claim Construction Order at 35. Defendant argues that each of these limitations is present in the “host kernel” component of each Y-System workstation’s operating system. Specifically, defendant points to the “inter-kernel protocol,” which provides for inter-process communications among V-System workstations, as teaching a plurality of “peer-level facilities] that supports direct communication with other peer-level facilities.” See V-System Manual at 1-3; see also First McKusick Decl. ¶ 19. Defendant also submits undisputed evidence establishing that the inter-kernél protocol is capable of multi-tasking — i.e., it is able to manage two or more concurrent or interleaved tasks. First McKusick Decl. ¶ 19.

Nonetheless, as the court’s claim construction order makes clear, the “multitasking interface function” must be comprised of software that is tailored to the functions performed by each of the peer-level facilities. First Claim Construction Order at 35. Defendant contends that this claim limitation is taught by the “device servers” described in the V-System Manual. These servers manage the various low-level input-output devices such as printers or keyboards that can be attached to one of the V-System workstations. V-System Manual at 1-4, 31-3; First McKusick Decl. ¶ 20. In contrast to the other server facilities disclosed in the V-System reference, which are executed outside the kernel, each device server is a “pseudoprocess” embedded in the host kernel, its precise characteristics varying with the type of device that it manages. V-System Manual at 31-3; McKusick Decl. ¶ 20; Third McKusick Decl. ¶¶ 21-22. Thus, according to defendant, the V-System Manual teaches a “multi-tasking interface function” that is “tailored” to the workstation on which it resides.

The court agrees that the device server component of each workstation’s host kernel is “tailored” to that “facility.” However, that does not necessarily mean the V-System’s host kernel discloses the multitasking interface function of claim 1. To do so, the tailored portion of the host kernel must “support[ ] direct communication with other peer-level facilities.” In the V-System server, that function is performed by the inter-kernel protocol component of each workstation’s host kernel, which is a separate part of the kernel that is unrelated to the device management processes carried out by the device servers. Thus, while it may be true that the configuration of each of the host kernels that comprise the V-System’s “distributed kernel” depends upon the devices that are connected to ap particular workstation, there is no evidence that would suggest that the communications-related functions of the host kernels are tailored to the tasks performed by the workstations on which they reside.

In short, the fact that the device server components of the V-System operating system reside in the host kernels of the server’s constituent workstations is not sufficient to support a finding that the V-System meets the “tailored to each ... peer-level facility” limitation of the claimed multi-tasking interface function. Accordingly, given the absence of any other evidence that the communications-related functions of the V-System’s distributed kernel are tailored to the particular facility on which they reside, a reasonable jury would be compelled to conclude that the V-System Manual does not teach the “multi-tasking interface function” limitation of the patented invention. The court therefore holds as a matter of law that claim 1 of the ’037 Patent is not anticipated by the V-System reference.

While defendant concedes that none of other prior art references that it has identified anticipates claim 1 as a matter of law, it nevertheless asserts that the teachings of the Cheriton and Epoch Server references give rise to genuine issues of material fact that preclude entry of summary judgment in plaintiffs favor. Here too, the court finds the absence of the “multi-tasking interface function” claim element to be dispositive. In each of these references, defendant identifies a component of a multiprocessor operating system that provides for communication between the processors that comprise the operating system. See Second McKusick Decl. ¶ 9 (describing the “interprocess communication facility” taught by Cheriton); Second McKusick Decl. ¶ 16 (noting the processors of the Epoch server “necessarily communicate with each other using a software program”). However, defendant’s reliance on these references suffers from the same problem as its attempt to show that the V-System anticipates claim 1, as neither Cheriton nor the Epoch Server teaches interprocessor communications software that is “tailored” to the tasks performed by each of the operating system’s constituent processors. Furthermore, defendant fails to point to any evidence that would suggest that one of ordinary skill in the art would have recognized such tailoring of communications-related functionality as being inherently disclosed by these references. Thus, for reasons stated in the court’s discussion of the Y-System Manual, neither the Cheriton nor the Epoch Server references could support a finding that claim 1 of the ’037 Patent is anticipated.

With respect to the remaining claims at issue, claims 2 through 6 depend on claim 1 and thus necessarily share the multitasking interface function limitation that is absent from the cited prior art. Furthermore, claims 7 through 9 recite a “multitasking interface” and claims 11 through 13 include a “multi-tasking interface sub-component” element. The parties agree that these claim elements are substantially identical to the “multi-tasking interface function” of claim 1, and the court has construed them accordingly. The court therefore grants plaintiffs motion for summary judgment of validity with respect to each of these claims.

CONCLUSION

For the reasons stated above, defendant’s motion for summary judgment of invalidity with respect to claims 1 and 3 of the ’366 Patent and claims 7 through 11 and 12 through 15 of the ’918 Patent is GRANTED IN PART and DENIED IN PART. Likewise, plaintiffs motion for summary judgment of validity with respect to the asserted claims of the ’366 and ’918 Patents is GRANTED IN PART and DENIED IN PART. In addition, plaintiffs motion for summary judgment of validity with respect to claims 1 through 9 and 11 through 13 of the ’037 Patent is GRANTED, and defendant’s motion for summary judgment of invalidity with respect to those claims is DENIED.

IT IS SO ORDERED.

1 . Plaintiff acquired the patents in suit in June 2003, following Auspex's decision to seek bankruptcy protection and cease operations.

2 . As discussed in the court’s previous orders, an Ethernet data packet containing an NFS request is comprised of a number of protocols, or "layers,” each of which provides some information about the source or contents of the request contained therein.

3 . Plaintiff’s separate motion for summary judgment of "no inequitable conduct” in the prosecution of the '037 Patent will be addressed in a separate order.

4 . Defendant also points to a number of "pre-process[ing]” functions that a person of ordinary skill in the art would have recognized to be disclosed by the front-end processing unit of the Epoch Server in support of its position that the front-end processor teaches the "decoding” and “encoding” claim limitations. See First McKusick Deck ¶ 49. However, even assuming that those functions are disclosed by the Epoch Server reference, the pre-processing protocols that defendant identifies involve little more than examining requests from a data network and determining whether those requests are intended for further processing by Epoch Server's file processor. For the reasons stated above, such protocols could not reasonably be found to perform the function of translating data from one format to another, decoded data format. Accordingly, like the process of serial-to-parallel conversion, the identified pre-processing functions could not support a finding that the Epoch Server reference teaches the "decoding” and "encoding” limitations of claim 1.

5 . It appears more likely that the LNFS format could be found to be equivalent to any similarly "reliable” data format — that is, a data format that can guarantee delivery of a particular sequence of data within a certain time and in particular order. Nevertheless, the court need not reach that issue here.

6 . The court reaches a different conclusion here than with it does with respect to claim 3 of the '366 Patent, despite the fact that claim 3 must similarly include a processing unit for decoding NFS requests and encoding NFS reply messages that lacks a general purpose operating system. This difference is entirely consistent with the limitations that section 112 ¶ 6 imposes of the scope of a means-plus-function claim element. Accord Pennwalt Corp. v. Durand-Wayland, Inc., 833 F.2d 931, 934 (Fed.Cir.1987) (en banc) (noting that section 112 ¶ 6 "rules out the possibility that any and every means which performs the function specified in the claim literally satisfies that limitation”), cert. denied, 485 U.S. 961, 108 S.Ct. 1226, 99 L.Ed.2d 426 (1988), and cert. denied, 485 U.S. 1009, 108 S.Ct. 1474, 99 L.Ed.2d 703 (1988). With respect to claim 5, those limitations restrict the claim’s scope to cover only structures programmed to process all necessary NFS protocol layers and their equivalents. The cited prior art simply does not contemplate implementing such an algorithm on a machine that lacks a general purpose operating system.

7 . In light of this conclusion, the court need not consider defendant's argument that Figure 1 of the ’918 Patent anticipates claims 7 through 10 and claims 12 through 15.