Kilopass Technology, Inc. v. Sidense Corporation

Filing 272

ORDER GRANTING DEFENDANT'S MOTION FOR SUMMARY JUDGMENT OF NON-INFRINGEMENT 241 248 250 251 257 268 (Illston, Susan) (Filed on 8/16/2012)

Download PDF
1 2 3 4 5 IN THE UNITED STATES DISTRICT COURT 6 FOR THE NORTHERN DISTRICT OF CALIFORNIA 7 8 KILOPASS TECHNOLOGY INC., 9 United States District Court For the Northern District of California 10 11 12 No. C 10-02066 SI Plaintiff, ORDER GRANTING DEFENDANT’S MOTION FOR SUMMARY JUDGMENT OF NON-INFRINGEMENT v. SIDENSE CORPORATION, Defendant. / 13 14 Defendant Sidense Corp. filed a motion for summary judgment of non-infringement in this patent 15 case. A hearing was held on this matter on August 10, 2012. Having considered the papers submitted, 16 and for good cause shown, the Court GRANTS defendant’s motion for summary judgment. 17 18 BACKGROUND 19 Plaintiff Kilopass Technology, Inc., is a company that markets “a novel way of storing data 20 permanently inside integrated circuits . . . by creating a breakdown in the transistor, safely and reliably.” 21 Third Am. Compl. (“TAC”) ¶ 8. Kilopass alleges that defendant Sidense Corporation “has knowingly 22 copied Kilopass’ patented technology and has been selling and offering for sale Kilopass’ patented 23 technology without authorization from Kilopass.” TAC ¶ 18. In particular, Kilopass alleges that 24 defendant infringed U.S. Patent Nos. 6,940,751 (“’751 Patent”), 6,777,757 (“’757 Patent”), and 25 6,856,540 (“’540 Patent”) (collectively, the “patents-in-suit”). TAC ¶¶ 29, 38, 45. Kilopass also alleges 26 that Sidense has been “sowing deceit in the marketplace” by “falsely alleg[ing] in the marketplace 27 statements to the effect that Kilopass has no intellectual property issues with Sidense.” TAC ¶ 24. 28 Kilopass originally filed suit in this Court on May 14, 2010. Kilopass filed its Second Amended 1 Complaint on October 14, 2010. On December 13, 2010, the Court granted in part and denied in part 2 a motion to dismiss, leaving six causes of action in the case: three patent infringement claims, one false 3 advertisement and disparagement claim, one intentional interference with prospective economic 4 relations claim, and one unfair competition claim. On August 31, 2012, the Court issued its Markman 5 Claim Construction Order. On May 22, 2012, Kilopass filed the Third Amended Complaint.1 6 1. 8 The patents-in-suit all generally relate to memory devices and arrays of such devices. Hutchins 9 Decl., Ex. 10 (Neikirk Expert Report), ¶ 10. Specifically, they relate to non-volatile memory devices, 10 United States District Court For the Northern District of California 7 which are memory devices that retain their state even when power is removed. Id. The programmable 11 memory cell is comprised of a transistor located at the cross point of a column bitline and a row 12 wordline. See ‘751, Abstract. The transistor has a gate formed from the column bitline and a source 13 connected to the row wordline. Id.2 There is a substrate underlying the gate. The memory cell is 14 programmed by applying a voltage potential between the column bitline and the row wordline to 15 produce a “programmed n+ region” in the substrate. Id. The key to programming the cell is “breaking 16 down” the gate dielectric into a damaged state. Breaking down the gate oxide requires applying a 17 sufficient voltage to the gate. In an undamaged state, during a read operation, very little current will 18 be able to flow from the gate into the channel region and then out to the device; this will provide 19 knowledge to the device that the cell is in its first binary state. Neikirk Exp. Rep. ¶ 10. In the damaged 20 state, current flow will be easier during the reading operation, and will provide knowledge that the cell 21 is in its second binary state. Id. 22 The Patents-in-Suit Figure 3 of the patents-in-suit depicts the memory cell: 23 24 25 26 27 28 1 On August 19, 2011, Sidense filed suit against Kilopass in a related case, Sidense v. Kilopass, CV 11-4112 SI, asserting various business torts against Kilopass, including defamation, false advertising, and intentional interference with contractual relations. 2 The use of only two electrical connections (one to the gate and one to the source) differentiates the patents-in-suit from other metal-oxide semiconductor field effect transistors (MOSFETs) which have three outside electrical connections: one to the gate, one to the source, and one to the drain. Neikirk Ex. Rep. ¶ 31. The use of only two connections rather than three enhances the density of the memory device, making it smaller. Neikirk Exp. Rep. ¶ 35. 2 1 2 3 4 5 6 7 8 9 United States District Court For the Northern District of California 10 ’757, Fig. 3. R1 depicts the row wordline (connected to the source) and C1 depicts the column bitline 11 (connected to gate). The gate structure is demarcated as 310 and the source region is 306. As noted, 12 the gate oxide layer (304) is broken down by the voltage potential during programming. 13 For the sake of example, in one embodiment described in the patents-in-suit, the voltage on a 14 selected row wordline is zero volts and the voltage on the column bitline is 8 volts. ’757, 6:22-24. 15 Thus, the voltage across the gate (which is connected to the bitline) of the transistor and the source 16 (connected to the wordline) is 8 volts. The gate oxide of the transistor is designed to break down at this 17 potential difference, which programs the cell. Id. “During programming, the voltage potential breaks 18 down the gate oxide and results in a leakage current through the gate oxide into the underlying substrate 19 and collected mostly by the N+ source/drain, which is connected to the ground.” Id., 6:30-34. In this 20 broken down state, during a later read operation (using 1.8 volts), current will flow from the column 21 bitline, through the gate oxide of the transistor (104), and out through the wordline – indicating the cell 22 is programmed. Id., 8:1-6. If, however, during programming, the voltage on the wordline is zero volts 23 and the voltage on the bitline is 3.3 volts, the potential difference of 3.3 volts is insufficient to break 24 down the gate oxide for the transistor at the crosspoint, and the memory cell does not program. Id., 25 6:51-55. During a read operation, because the gate oxide is not broken down, no current will flow, 26 which will indicate to the device that the memory cell is not programmed. Id, 8:2-12. 27 28 3 1 2. 2 Defendant Sidense Corporation is active in “the emerging market for ‘one transistor’ (1T bitcell), 3 one-time programmable (OTP) embedded non-volatile memory (‘eNVM’).” Sidense v. Kilopass, CV 4 11-4112, SI, First Amend. Compl. (“FAC”) ¶ 7. Sidense’s product is called the 1T-Fuse, and is used 5 by its customers in a “variety of integrated circuits, which are then placed in a consumer electronics 6 product such as a smartphone or set top box.” Id., ¶ 7. Sidense does not make, use, or sell any apparatus 7 or method; rather, it licenses the design of its technology to customers, who in turn use those designs 8 to build embedded bit cells and bit cell arrays in chips manufactured by or for them. See Def.’s Mot. 9 to Dismiss at 11 (Dkt. 12). Sidense’s only product is its memory macro-cell design, which Sidense United States District Court For the Northern District of California 10 11 The Alleged Infringing Technology licenses to customers who download “GDS” files containing those designs from its servers. Id. Sidense’s technology was invented by Wlodek Kurjanowicz. Hutchins Decl., Ex. 1 12 (Kurjanowicz Dep.), 34:2-5. Kurjanowicz testifies that in an earlier memory circuit he had worked on, 13 gate oxide breakdown was a problem. Id., 32:18-22. In 2003, he realized that in fact gate oxide 14 breakdown could be useful, and “tried to use that knowledge to develop memory that would use [gate 15 oxide breakdown] as a programming phenomenon.” Id. He began to work on developing memory cells 16 that utilize the breakdown. However, on the night of February 4, 2004, his partner informed him about 17 Kilopass and its one-time programmable memory patents. In reviewing the patents, Kurjanowicz 18 recognized that Kilopass had disclosed 1.5T cells that “looked very similar or if not identical to what 19 [he] was designing for last many months.” Id., 35:16-18. He then undertook a new design, which he 20 recorded in his notebook that night, of a “half transistor” anti-fuse memory design. Id., 34:2-5; Def.’s 21 Opp. at 7. This is the technology that Sidense uses to this day.3 Def.’s Opp. at 7. Sidense has received 22 16 patents on its technology, including a patent for its “Split Channel Antifuse Array Architecture,” 23 Patent No. 7,402,855 (the “‘855 Patent”). 24 Kilopass seeks to establish infringement based on a chip it received from one of Sidense’s 25 customers, XMOS. The chip includes Sidense’s 1T OTP technology. Neikirk Exp. Rep. ¶ 38. Kilopass 26 provided the chip to a third company to “decap” the chip, which involved cutting the chip and 27 3 28 The product Sidense markets as the 1T-Fuse is the “half a T memory architecture” designed by Kurjanowicz. Id., 34:2-5. 4 1 examining it under a Transmission Electron Microscope (“TEM”). The images provided from that 2 analysis form the basis of Kilopass’ expert’s analysis of Sidense’s 1T OTP memory structure. Kilopass 3 also points to aspects of Sidense’s ‘855 patent to demonstrate infringement. 4 5 3. 6 Kilopass asserts claims 1, 2, 7, 8, and 14 of the ‘757 patent; claims 1, 3, 5, 6, 9, 12, and 14 of 7 the ‘751 patent; and claims 1 and 5 of the ‘540 patent. Claim 1 of ‘751 claims the following (terms 8 relevant to Sidense’s motion in bold): 9 United States District Court For the Northern District of California 10 11 12 13 14 15 The Asserted Claims 1. A programmable memory cell useful in a memory array having column bitlines and row wordlines, the memory cell comprising: a transistor having a gate, a gate dielectric between the gate and over a substrate, and first and second doped semiconductor regions formed in said substrate adjacent said gate and in a spaced apart relationship to define a channel region therebetween and under said gate; and wherein the second doped semiconductor region of the transistor is connected to one of said row wordlines, and wherein said gate dielectric is formed such that the gate dielectric is more susceptible to breakdown near the first doped semiconductor region than said second doped semiconductor region. 16 ‘751, 14:29-44. Sidense argues that four of the limitations described in this claim, which are shared by 17 all of the asserted claims, are not found in its technology: (1) a “row wordline” connected to the “second 18 doped semiconductor region”; (2) a “transistor”; and (3) “first and second doped semiconductor regions” 19 that are (4) “in a spaced apart relationship.” Def.’s Mot. at 9. 20 Sidense moved for summary judgment of non-infringement on July 11, 2012. Sidense only 21 moves on the patent claims, not Kilopass’ tort claims. Kilopass filed an opposition to the motion for 22 summary judgment on July 20, 2012. Dkt. 250. Along with its motion for summary judgment, Sidense 23 filed a Daubert motion to exclude the testimony of Kilopass’ expert Neikirk. Dkt. 248. For its part, 24 Kilopass filed an objection to the declaration of Sidense’s expert filed along with Sidense’s motion. 25 26 27 28 5 1 Dkt. 253.4 Kilopass also filed a motion to strike Sidense’s reply. Dkt. 268.5 2 3 LEGAL STANDARD Summary adjudication is proper when “the movant shows that there is no genuine dispute as to 5 any material fact and the movant is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(a). In a 6 motion for summary judgment, “[if] the moving party for summary judgment meets its initial burden 7 of identifying for the court those portions of the materials on file that it believes demonstrate the absence 8 of any genuine issues of material fact, the burden of production then shifts so that the nonmoving party 9 must set forth, by affidavit or as otherwise provided in Rule 56, specific facts showing that there is a 10 United States District Court For the Northern District of California 4 genuine issue for trial.” See T.W. Elec. Service, Inc., v. Pac. Elec. Contractors Ass'n, 809 F.2d 626, 630 11 (9th Cir.1987) (citing Celotex Corp. v. Catrett, 477 U.S. 317, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986)). 12 In judging evidence at the summary judgment stage, the Court does not make credibility determinations 13 or weigh conflicting evidence, and draws all inferences in the light most favorable to the non-moving 14 party. Id. at 630–31 (citing Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 106 15 S.Ct. 1348, 89 L.Ed.2d 538 (1986)). The evidence presented by the parties must be admissible. See 16 Fed.R.Civ.P. 56(c)(4). Conclusory, speculative testimony in affidavits and moving papers is insufficient 17 to raise genuine issues of fact and defeat summary judgment. See Thornhill Publ'g Co., Inc. v. GTE 18 Corp., 594 F.2d 730, 738 (9th Cir.1979). 19 To find infringement, “the court must determine that every claim limitation is found in the 20 accused device.” Playtex Products, Inc. v. Procter & Gamble Co., 400 F.3d 901, 909 (Fed. Cir. 2005) 21 (internal citations omitted). Summary judgment of non-infringement is a two-step analysis. First, the 22 claims of the patent must be construed to determine their scope, as a question of law. Pitney Bowes, Inc. 23 v. Hewlett-Packard Co., 182 F.3d 1298, 1304 (Fed. Cir. 1999) (internal citation omitted). Second, "a 24 determination must be made as to whether the properly construed claims read on the accused device." 25 26 4 27 5 28 Because the Court does not rely on that declaration, it need not address this objection. Kilopass moves to strike Sidense’s reply on the grounds that it was filed 38 minutes late and presents new argument with respect to the rowline/wordline limitation. The Court does not rely on Sidense’s new argument in this Order. The Court otherwise DENIES the motion to strike. 6 1 Id. The determination of infringement is generally a question of fact. Lockheed Martin Corp. v. Space 2 Sys./Loral, Inc., 324 F.3d 1308, 1318 (Fed. Cir. 2003). Since the ultimate burden of proving 3 infringement rests with the patentee, an accused infringer may establish that summary judgment is 4 proper “either by providing evidence that would preclude a finding of infringement, or by showing that 5 the evidence on file fails to establish a material issue of fact essential to the patentee's case.” Novartis 6 Corp. v. Ben Venue Labs., Inc., 271 F.3d 1043, 1046 (Fed. Cir. 2001). 7 8 9 DISCUSSION United States District Court For the Northern District of California 10 As noted above, Sidense argues that four of the limitations set forth in all of the asserted claims 11 are not present in its technology: (1) “row wordlines” connected to the “second doped semiconductor 12 region”; (2) a “transistor”; and (3) “first and second doped semiconductor regions” that are (4) “in a 13 spaced apart relationship.” Def.’s Mot. at 1. Sidense argues that no question of material fact exists with 14 respect to any of the four limitations. Because a finding of infringement requires that every claim 15 limitation is found in the accused device, the lack of a question of material fact on any of the four 16 limitations warrants summary judgment in this case. See Playtex Products, Inc. v. Procter & Gamble 17 Co., 400 F.3d 901, 909 (Fed. Cir. 2005). 18 19 1. 20 All claims of the ‘757 and ‘540 patents have the limitation of “a row wordline segment coupled 21 to the second doped semiconductor region.” ‘757, 10:46-47 and 11:15-16; ‘540, 11:6-7. Sidense argues 22 that its technology does not meet this limitation because its memory cell has a row wordline connected 23 to the gate and a column bitline connected to the second doped semiconductor region, the opposite 24 configuration of Kilopass’ patents-in-suit. Sidense argues that this is undisputed, and points out that 25 Kilopass’ infringement expert cites to the following illustration of Sidense’s technology: “Row Wordlines” Connected to the “Second Doped Semiconductor Region” 26 27 28 7 1 2 3 4 5 6 7 Neikirk Exp. Rep. ¶ 69. In the illustration, Sidense’s wordline [WL] is connected to the gate, while the 8 bitline [BL] is attached to the second doped semiconductor region (the N+ diffusion region). See Pl.’s 9 Mot. at 17. As noted, this is the opposite configuration from diagrams disclosed in Kilopass’ patents-in- United States District Court For the Northern District of California 10 suit, which connect the gate to the bitline and the second doped semiconductor to the wordline: 11 12 13 14 15 16 17 ‘751, Fig. 27. 18 Recognizing its opposite configuration, Kilopass originally argued at claim construction that 19 wordlines and bitlines were interchangeable terms. See Kilopass’s Opening Claim Const. Br. at 5 20 (“[O]ne of ordinary skill in the art would understand that the current flows can be detected in both the 21 bitline and the wordline, again showing the interchangeability of the two, and with the naming of 22 ‘bitline’ or ‘wordline’ being simply a matter of perspective.”). As interchangeable terms, Kilopass 23 proposed they be defined identically as “a line that connects to one terminal of each memory cell in a 24 memory array.” Sidense argued that a bitline was limited to a “line that connects the memory cell to 25 the sensing circuit during the read operation,” while a wordline is a “line connected to the memory cell, 26 which is selected by the row addresses.” Claim Const. Order at 8. 27 The Court did not adopt Kilopass’s construction, but largely found in its favor. The Court noted 28 that wordlines and bitlines always appeared orthogonal to one another in a memory array, and thus “the 8 1 Court will not define two different terms to mean precisely the same thing when they are not identical.” 2 Claim Const. Order at 9. However, the Court limited the differences between bitlines and wordlines to 3 their positions in relation to one another: it defined bitline/column bitline as “a line orthogonal to the 4 row wordline that connects to a terminal of each memory cell in a memory array,” and wordline/row 5 wordline as “a line orthogonal to the column bitline that connects to a terminal of each memory cell in 6 a memory array.” Id. However, during inter partes reexamination proceedings of its own patents before the Patent & 8 Trademark Office, Kilopass adopted a contradictory position from its earlier argument before this Court 9 in order to overcome prior art. See May 1, 2012 Order Upon Reconsideration, Dkt. 224, at 5. Sidense, 10 United States District Court For the Northern District of California 7 which instigated the reexamination, had argued to the PTO that the Kilopass’ ‘751 patent was 11 anticipated by an earlier patent, Tanaka et al. (U.S. Patent No. 5,331,181) (“Tanaka”). In Tanaka, unlike 12 Kilopass’s ‘751 patent but like Sidense’s cell, the doped semiconductor region is connected to a bitline. 13 The patent examiner at the PTO proceeding ruled that Kilopass overcame Tanaka because “it is well 14 known to one of ordinary skill in the art at the time of the invention that the bitlines and wordlines have 15 distinct functional effect on the operation of memory devices and thus are not interchangeable.” See 16 Khaliq Decl., Ex. 4 (Feb. 18, 2011 USPTO Office Action). After Sidense appealed that decision to the 17 PTO’s Board of Patent Appeals and Interferences (the “BPAI”), Kilopass filed a brief explicitly 18 agreeing with the Patent Examiner’s finding: 19 20 21 22 23 24 25 26 27 28 With respect to claims 5 and 11, the Patent Owner agrees with the Examiner that Tanaka does not show a gate formed from a column bit line. As can been seen [sic] in Figure 2(b) of Tanaka, the gates of the transistors are coupled to row wordlines. Therefore claims 5 and 11 are not anticipated by Tanaka. See Hutchins Decl., Ex. 6 at 8 (Kilopass’s Jan. 6, 2012 BPAI Brief). The position Kilopass took before the BPAI was clearly irreconcilable with its “interchangeability” position before this Court. In a May 1, 2012 Order, this Court found that by taking the contrary position it did before the BPAI, Kilopass clearly and unmistakably disavowed claim scope where gates of transistors are connected to row wordlines. See Dkt. 224 at 9 (citing Computer Docking Station Corp. v. Dell, Inc., 519 F.3d 1366, 1374 (Fed. Cir. 2008) (noting that a patentee can disavow claim scope “by clearly characterizing the invention in a way to try to overcome rejections based on 9 1 prior art”); Spectrum Intern., Inc. v. Sterilite Corp., 164 F.3d 1372, 1379 (Fed. Cir. 1998) (“Claims may 2 not be construed one way in order to obtain their allowance and in a different way against accused 3 infringers.”)). Kilopass has therefore disavowed claim scope for transistors with gates connected to wordlines, 5 and the figures relied on by Kilopass’s expert illustrate that Sidense’s gates are connected to wordlines.6 6 To avoid summary judgment on this issue, Kilopass has introduced a new theory. In his June 1, 2012 7 rebuttal expert report, Neikirk argues that Sidense’s wordlines – the lines connected to the gates – 8 actually “perform the identical function of the bitlines of the Asserted Patents, while the lines connected 9 to the sources in Sidense products perform the identical function of the wordlines of the Asserted 10 United States District Court For the Northern District of California 4 Patents.” Neikirk Reb. Rep. ¶ 42. According to Neikirk, the “function” of a bitline in the patents-in-suit 11 is to “supply a higher voltage than that supplied to the source terminal via the line that is orthogonal to 12 the line connected to the gate terminal (wordline). This creates a voltage potential across the gate 13 terminal and the source terminal (a higher voltage at the gate and ground or lower voltage at the 14 source)” which allows for a breakdown of the gate oxide. Id. ¶ 78. Because Sidense’s “row wordline” 15 provides for the higher voltage, according to this theory, it is de facto a bitline. And because it is a 16 bitline, it is not affected by the Court’s disavowal order. Id.¶¶ 88-89. 17 One problem with Kilopass’s new theory, however, is that it is new. Kilopass did not argue a 18 function/voltage theory of bitlines and wordlines during claim construction, let alone prevail on such 19 an argument. Instead, Kilopass argued that the two terms should be given the same definition. See 20 Kilopass’s Opening Claim Const. Br. at 4-5. “Once a district court has construed the relevant claim 21 terms, and unless altered by the district court, then that legal determination governs for purposes of trial. 22 No party may contradict the court's construction to a jury.” Exergen Corp. v. Wal-Mart Stores, Inc., 575 23 F.3d 1312, 1321 (Fed. Cir. 2009). Kilopass cannot now successfully argue that a bitline is defined as 24 the higher voltage line. 25 Moreover, Kilopass’s theory would render the May 1, 2012 disavowal Order meaningless. In 26 27 28 6 A February 7, 2008 internal email from Kilopass to its patent counsel at Perkins Coie notes that Sidense has “changed the bit line with word line in the split gate bit cell since apparently we did not manage our patents in the beginning as well as we might have done.” Hutchins Decl., Ex. 7. 10 the preferred embodiment described in Tanaka, the line connected to the gate (Tanaka’s “wordline”) is 2 set to 20 volts, and the line connected to the source (Tanaka’s “bitline”) is set to zero volts. Hutchins 3 Decl., Ex. 18 (U.S. Patent No. 5,331,181), 3:66-4:9. Under Kilopass’s voltage theory, the Tanaka 4 wordline, as the line providing higher voltage, would in fact be the bitline. However, Kilopass argued 5 that its patents were not anticipated by Tanaka because Tanaka connected the gates to row wordlines. 6 See Hutchins Decl., Ex. 6 at 8 (Kilopass’s Jan. 6, 2012 BPAI Brief) (“As can been seen [sic] in Figure 7 2(b) of Tanaka, the gates of the transistors are coupled to row wordlines. Therefore claims 5 and 11 are 8 not anticipated by Tanaka.”). To accept Kilopass’s definition of bitlines as simply the lines that provide 9 higher voltage would also recapture the Tanaka configuration, which Kilopass expressly disclaimed. 10 United States District Court For the Northern District of California 1 The Court finds that there is no genuine issue of material fact that Sidense connects its gate to 11 a row wordline, and that Kilopass expressly disavowed that claim scope from the patents-in-suit. The 12 “row wordline segment coupled to the second doped semiconductor region” limitation is not met. 13 14 2. 15 All the of the asserted claims describe a “transistor” having a “gate, a gate dielectric between 16 the gate and over a substrate, and first and second doped semiconductor regions” formed in the 17 substrate. ‘751, 14:34-35; ‘540, 10:67-11:1 (emphasis added). The parties agree that Sidense’s memory 18 cell has only one “doped semiconductor region,” the “second doped semiconductor region.” The parties 19 also agree that Sidense does not have a “first doped semiconductor region” and thus this limitation is 20 not literally infringed. Kilopass argues, however, that the shallow trench isolation (“STI”) insulator 21 contained in Sidense’s memory cell is the equivalent of a first doped semiconductor region under the 22 doctrine of equivalents. “First Doped Semiconductor Region” 23 The doctrine of equivalents holds that even if an accused product does not literally infringe the 24 asserted claims of a patent, the product may infringe if the differences between the element of the 25 accused product at issue and the claim limitation at issue are insubstantial. Dawn Equip. Co. v. 26 Kentucky Farms, Inc., 140 F.3d 1009, 1015-16 (Fed.Cir.1998). To oppose a defendant's motion for 27 summary judgment of non-infringement under the doctrine of equivalents, the plaintiff has the burden 28 of producing “particularized testimony and linking argument on a limitation-by-limitation basis that 11 1 create[s] a genuine issue of material fact as to equivalents.” AquaTex Indus., Inc. v. Techniche 2 Solutions, 479 F.3d 1320, 1328–29 (Fed.Cir.2007). Whether equivalency exists may be determined 3 based on the “insubstantial differences” test or based on the “function-way-result,” which asks whether 4 the element of the accused device “performs substantially the same function in substantially the same 5 way to obtain the same result.” TIP Systems, LLC v. Phillips & Brooks/Gladwin, Inc., 529 F.3d 1364, 6 1376-77 (Fed. Cir. 2008) (citing Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 38-40 7 (1997)). As an initial matter, Sidense argues – and the Court agrees – that Kilopass amended its 9 infringement contentions with respect to the doctrine of equivalents far past the applicable deadlines 10 United States District Court For the Northern District of California 8 without Court approval. This District has adopted Patent Local Rules that “require parties to state early 11 in the litigation and with specificity their contentions with respect to infringement and invalidity.” O2 12 Micro Int'l, Ltd. v. Monolithic Power Sys., Inc., 467 F.3d 1355, 1359 (Fed.Cir.2006).7 Kilopass served 13 its Amended Infringement Contentions pursuant to Patent Local Rule 3-1 and 3-2 on April 4, 2011. 14 Hutchins Decl., Ex. 21 (Ex. 3). In those contentions, Kilopass asserted the following “function, way, 15 result” theory: 16 Additionally, the STI is the equivalent of the first doped region. The function of the first doped region is to provide a channel stop. The way it functions is to prevent current from the channel to flow in the area of the STI. The result is that the end of the channel is defined. Sidense’s STI performs substantially the same function, functions in substantially the same way, and achieves the substantially same result. 17 18 19 20 21 22 23 24 Id. at 3. Sidense provides evidence that it expended resources to prepare a defense to this theory, including deposing the named inventors on the patents-in-suit as well as Kilopass’ CTO and VP of Engineering on issues relating to “channel stop.” Giardina Decl. ¶ 10. A year later, on April 13, 2012, Kilopass filed its expert report from Dr. Neikirk with a different equivalency theory. In that report, Neikirk asserts that: The function of the first and second regions is clearly stated: to geometrically delineate or define a channel region between them . . . The way in which each region functions to geometrically delineate or define a channel region is by being 25 26 27 28 7 Patent Local Rule 3-1 provides that, “Not later than 14 days after the Initial Case Management Conference, a party claiming patent infringement shall serve on all parties a ‘Disclosure of Asserted Claims and Infringement Contentions.’” 12 1 separated from each other in a cross section of the device, by being in the substrate, and by being adjacent to the gate . . . The result produced by the two regions required by this claim limitation is to create or delineate a defined region of the device that is a channel region. 2 3 Id. ¶¶ 62-64. This equivalence theory focuses on the geometry of the cell, rather than the electrical 4 properties of the “channel stop.” Kilopass argues that its new theory does not “substantially depart” and 5 is “consistent” with its original theory. Pl.’s Opp. to Def.’s Daubert Mot. at 9-10. However, the Patent 6 Local Rules allow amendment of infringement contentions “only by order of the Court upon a timely 7 showing of good cause.” P.L.R. 3-6. It is for the Court to make the determination of whether a new 8 9 United States District Court For the Northern District of California 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 theory is consistent with an earlier contention or otherwise appropriate, not the party asserting it.8 See Spectros Corp. v. Thermo Fisher Sci., 2012 WL 1965887, *6 (N.D. Cal. May 31, 2012) (Armstrong, J.) (procedurally barring doctrine of equivalents theory for failure to amend infringement contentions). The failure to provide and support an admissible equivalents theory in the opposition papers warrants summary judgment on the “first doped semiconductor region” limitation. Moreover, Kilopass’s doctrine of equivalence theory fails on the merits. It is undisputed that the STI is an insulator. See Pl.’s Opp. at 18 (referring to an “STI Insulator”). It is also undisputed that the “first doped semiconductor region” is a semiconductor. At claim construction, the parties agreed that a “semiconductor” is “a material, like silicon, whose conductivity is in the range between that of metals and insulators, and whose conductivity can be altered by the introduction of an impurity.” See Claim Const. Order at 14. By the definition set forth at Claim Construction, a semiconductor is not an 8 Kilopass’s assertion of a new theory of equivalence is particularly inappropriate in light of evidence that Kilopass has known for many years that Sidense does not literally infringe its patents. On January 20, 2006, Kilopass’ patent counsel at Perkins Coie emailed Kilopass regarding Sidense’s technology: “[W]e speculated that Sidense may have eliminated the first doped region and replaced it with a shallow trench isolation of some sort. While possible, it is interesting that they have emphasized the need for the floating diffusion doped region as being advantageous to their design . . . Still, if in fact they have eliminated the first doped region, then they would NOT infringe our claims literally. If that is the case, then we would have to go through a ‘reissue’ proceeding in the patent office that may take 2 years in order to modify our claims to include the situation where there is no first doped region.” Hutchins Decl., Ex. 6 at 2. (emphasis in original). 28 13 1 insulator, nor its equivalent. Kilopass has provided no evidence that insulators are the equivalent of 2 semiconductors. Instead, the evidence shows the contrary. See Gosney Ex. Rep. at ¶ 151 (“[I]nsulators 3 have substantially different characteristics from semiconductors”). The uncontroverted evidence 4 demonstrates that there is a non-trivial difference between insulators and semiconductors. See Festo 5 Corp. v. Shoketsu Kogyo Kabushiki Co., 535 U.S. 722 (2002) (stating that “[t]he doctrine of equivalents 6 allows the patentee to claim those insubstantial alterations that were not captured in drafting the original 7 patent claim but which could be created through trivial changes.”) (emphasis added). Instead of arguing that insulators are equivalent to semiconductors, Kilopass’s expert argues that 9 the electrical properties of the “first doped semiconductor region” are irrelevant; he argues that “the 10 United States District Court For the Northern District of California 8 function required by this claim element is only geometrical/structural.” Neikirk Reb. Rep. ¶ 68. 11 However, this would render the term “semiconductor” – a term defined with respect to its electrical 12 conductivity – superfluous. This would be improper under Federal Circuit precedent. See Bicon, Inc. 13 v. Straumann Co., 441 F.3d 945, 950 (Fed. Cir. 2006) (“Allowing a patentee to argue that physical 14 structures and characteristics specifically described in a claim are merely superfluous would render the 15 scope of the patent ambiguous, leaving examiners and the public to guess about which claim language 16 the drafter deems necessary to his claimed invention and which language is merely superfluous, 17 nonlimiting elaboration.”). The Court will not ignore the actual terms in the claim element. 18 Kilopass also ignores the not-insubstantial difference that its use of a first doped semiconductor 19 region, rather than an STI, renders its memory cell smaller. In the specifications of the patents-in-suit, 20 Kilopass distinguishes prior art by stating that “in each of the memory cells described above, the cell 21 size is relatively large. The present invention provides a much smaller cell size, thereby allowing a 22 higher density.” ‘757, 4:45-47. Kilopass does not dispute that use of a “first doped semiconductor 23 region” allows it to achieve higher density by combining the first doped semiconductor regions of 24 adjacent cells rather than use STI regions. At deposition, the lead inventor of the patents-in-suit, Jack 25 Peng, stated that the reason for having two diffusions per cell was “because we want to do this to make 26 it a higher density . . .” Hutchins Decl., Ex. 20 (Peng Dep.), 133:7-11. In an email from Peng to 27 Kilopass’ patent counsel, Peng stated that the reason Kilopass did not implement Sidense’s cell design 28 “in our product, [is] because this split gate cell is not self-aligned, so their practical cell size will be 14 1 larger than our 1.5T cell.” Hutchins Decl, Ex.3 (Nov. 19, 2005 Peng email).9 In the nanometer world 2 of memory cells, a feature allowing for higher density and smaller size is more than an ancillary benefit; 3 it is one of the central purposes of the design. The use of the first doped semiconductor over an STI 4 therefore represents more than an insubstantial difference. Kilopass attempts to avoid the numerous differences between its first doped semiconductor 6 region and the STI insulator by proposing a broad “function-way-result” of the claim element. As noted 7 above, Kilopass argues that the “function” of the first doped semiconductor region is to “geometrically 8 delineate or define a channel region between” it and the second doped semiconductor region; the “way” 9 is “by being separated from each other in a cross section of the device”; and the “result” produced is “to 10 United States District Court For the Northern District of California 5 create or delineate a defined region of the device that is a channel region.” Neikirk Exp. Rep. ¶¶ 62-64. 11 In Instituform Technologies, Inc. v. Cat Contracting, Inc., 161 F.3d 668 (Fed. Cir. 1998), the Federal 12 Circuit reversed a District Court’s finding of infringement under the doctrine of equivalents where the 13 plaintiff argued that defendant’s use of a needle was the same as plaintiff’s use of a cup in a method for 14 creating a vacuum. Id. at 690. The District Court agreed with the plaintiff that there was equivalence 15 after determining the function of the cup was to “suck a vacuum,” the “way” as “being connected to a 16 vacuum source,” and the “result” being to “produce a satisfactory vacuum at the resin front.” Id. The 17 Federal Circuit reversed the finding of infringement, by “hold[ing] that the district court’s function-way- 18 result analysis involves too much overlapping and is overly broad.’” Id. at 693-94. Similarly here, 19 Kilopass seeks to avoid the numerous differences between its “first doped semiconductor region” and 20 the STI insulator by simply providing an overly broad function-way-result of the claim element, a 21 definition that focuses on the fact that the “first doped semiconductor region” is separated from the 22 second region, but ignores the “way” of using semiconductive (rather than insulative) material, and the 23 “result” that it produces a higher density memory array. As the Federal Circuit held in Instituform, a 24 plaintiff’s broad, overlapping definition of the function-way-result of a claim term cannot vitiate the 25 differences between the claim and the allegedly infringing technology. Instituform, 161 F.3d at 693-94. 26 9 27 28 The email references the fact that the patents-in-suit use a first doped semiconductor region in a “self-aligned” design, whereas Sidense’s STI region is “masked-aligned.” Gosney Ex. Rep. ¶ 164. The fact that Kilopass uses a self-alignment process rather than a mask-aligned process represents another difference between the two cells. See Gosney Decl. ¶ 158. 15 1 Summary judgment is therefore warranted on the claim limitation of a “first doped 2 semiconductor region,” which is neither present in Sidense’s memory cell nor equivalent to the STI 3 insulator. 4 3. 6 The asserted claims require that the first and second doped semiconductor regions be “in a 7 spaced apart relationship.” ‘751, 14:35-37. Assuming for the purposes of this discussion that an STI 8 insulator is equivalent to the first doped semiconductor region, the question is whether the STI and the 9 second doped semiconductor region are in “a spaced apart relationship.” At claim construction, the 10 United States District Court For the Northern District of California 5 parties agreed that “a spaced apart relationship” means “not in physical contact.” Kilopass does not 11 dispute the fact that the STI and the second doped semiconductor region are, in fact, in physical contact. 12 Kilopass argues that they nonetheless meet the “spaced apart” limitation because they are not in physical 13 contact in a cross-section of the cell. Kilopass points to the space between the STI and the second doped 14 semiconductor region, which forms the channel region. For its part, Sidense provides a diagram of a 15 top down view of the cell, which clearly shows the STI in physical contact with the second doped 16 semiconductor region. See Gosney Exp. Rep. ¶ 140. “Spaced Apart Relationship” 17 Kilopass argues that “nowhere in the Asserted Patents is the three dimensional (3D) construct 18 of a bitcell mentioned or described. Particularly, none of the elements in a bitcell structure is defined 19 in the context of other than a 2D cross section.” Pl.’s Opp. at 20. However, Kilopass’s construction 20 would require the Court to import a new limitation into the claim language: first and second doped 21 semiconductor regions “in a spaced apart relationship in a cross section.” This would be an improper 22 construction of the claim language. See Hoganas AB v. Dresser Indus., Inc., 9 F.3d 948, 950 23 (Fed.Cir.1993) (“It is improper for a court to add extraneous limitations to a claim, that is limitations 24 added wholly apart from any need to interpret what the patentee meant by particular words or phrases 25 in the claim.”). It would also render the “spaced apart” requirement superfluous, because first and 26 second doped semiconductor regions viewed from the side would always be spaced apart under to the 27 claim language: the first and second doped semiconductor regions are . . . “in a spaced apart relationship 28 to define a channel region therebetween . . .” ‘757, 10:43-44 (emphasis added). Kilopass’s definition 16 1 would “ascibe no meaning[] to the term . . . not already implicit in the rest of the claim.” Mangosoft Inc. 2 v. Oracle Corp., 525 F.3d 1327, 1330-31 (Fed. Cir. 2008). 3 4 The Court therefore finds summary judgment is warranted on the limitation that the first and second doped semiconductor regions be “in a spaced apart relationship.”10 5 6 CONCLUSION The Court finds that there is no genuine dispute of material fact that Sidense’s technology does 8 not infringe three of the claim limitations of the patents-in-suit. The Court therefore GRANTS 9 Sidense’s motion for summary judgment. This Order does not dispose of Kilopass’ tort claims against 10 United States District Court For the Northern District of California 7 Sidense for trade libel and defamation, intentional interference with prospective economic relations, and 11 false advertising and disparagement. TAC ¶¶ 51-65. 12 13 Docket Nos. 241, 248, 250, 251, 257, 268. 14 15 IT IS SO ORDERED. 16 17 Dated: August 16, 2012 SUSAN ILLSTON United States District Judge 18 19 20 21 22 23 24 25 26 27 10 28 Having found summary judgment is warranted on three of the claim limitations, the Court does not address Sidense’s argument with respect to the “transistor” claim. 17

Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.

Why Is My Information Online?