United States District Court, N.D. California, San Jose Division
April 4, 2014
RADWARE, LTD.; RADWARE, INC., Plaintiffs,
A10 NETWORKS, INC., Defendant, Counterclaim-Defendants, Counterclaim-Plaintiff. RADWARE, LTD.; RADWARE, INC., Plaintiffs, Counterclaim-Defendants,
F5 NETWORKS, INC., Defendant, Counterclaim-Plaintiff.
ORDER GRANTING PLAINTIFFS' MOTIONS FOR LEAVE TO AMEND INFRINGEMENT CONTENTIONS
HOWARD R. LLOYD, Magistrate Judge.
Plaintiffs Radware, Ltd. and Radware, Inc. (collectively, "Radware") sue defendants A10 Networks, Inc. ("A10") and F5 Networks, Inc. ("F5") (collectively, "Defendants") in two separate but related cases alleging that Defendants' products infringe three Radware patents related to "Load Balancing" technology. The related cases are subject to the same case management schedule, at least through claim construction. Radware served its preliminary infringement contentions ("PICs") in August 2013. Defendants made source code available for inspection by October 1, 2013. In mid to late December, Radware served its proposed amended infringement contentions ("AICs"), which amendments are purportedly based entirely on information gleaned from Defendants' productions of source code. Radware then filed the instant motions for leave to amend on January 28, 2014. Defendants oppose the motions. The matters were deemed suitable for determination without oral argument, and the March 21, 2014, hearing was vacated. See Civil L.R. 7-1(b). Based on the moving and responding papers, Radware's motions are granted.
A10's production of source code on October 1, 2013, consisted of nearly 300 total versions of the two alleged infringing product series, each version consisting of millions of lines of code. Radware consistently reviewed the available source code from October 8 through October 18, and again from November 4 through November 15; it did not inspect the source code during the two-week period in between. Radware requested hard copies of select portions of the source code in mid-November, some of which it did not receive until December 18. On December 19, Radware served A10 with its AICs, which added: (1) citations to source code; (2) new citations to documents previously cited; and (3) new doctrine of equivalents (DOE) theories for four claim elements. The AICs also eliminated previous assertions of infringement of 10 claims.
F5 produced its source code for inspection on September 30, which comprised nearly nine gigabytes of source code - more than one million lines of code. Radware began its review of the source code on October 11. It requested hard copies on October 15, which F5 did not produce until November 2. On December 23, Radware served F5 with its AICs, which added the same three categories of information described above. The AICs also removed eight claims previously alleged to have been infringed.
"Amendment of the Infringement Contentions... may be made only by order of the Court upon a timely showing of good cause." Patent L.R. 3-6. The good cause inquiry "considers first whether the moving party was diligent in amending its contentions and then whether the non-moving party would suffer prejudice if the motion to amend were granted. Acer, Inc. v. Tech. Props. Ltd., No. 08-cv-00882JF (HRL), 2010 WL 3618687, at *3 (N.D. Cal. Sept. 10, 2010) (citing O2 Micro Int'l Ltd. v. Monolithic Power Sys., Inc., 467 F.3d 1355, 1366-68 (Fed. Cir. 2006)). "The burden is on the movant to establish diligence rather than on the opposing party to establish lack of diligence." Karl Storz Endoscopy-America, Inc. v. Stryker Corp., No. C09-00355, 2011 WL 5574807, at *1 (N.D. Cal. Nov. 16, 2011) (quoting O2 Micro, 467 F.3d at 1366). "However, even if the movant was arguably not diligent, the court retains discretion to grant leave to amend." Linex Techs., Inc. v. Hewlett-Packard Co., No. C13-159 CW, 2013 WL 5955548, at *1 (N.D. Cal. Nov. 6, 2013); see also Apple Inc. v. Samsung Electronics Co., No. CV 12-00630 LHK, 2012 WL 5632618 (N.D. Cal. Nov. 15, 2012) (granting leave to amend infringement contentions, even though court found plaintiff failed to establish diligence, because of lack of prejudice to the defendants). "In considering the party's diligence, the critical question is whether the party could have discovered the new information earlier had it acted with the requisite diligence." Apple, 2012 WL 5632618, at *2 (internal quotation marks omitted).
"The rules are designed to require parties to crystallize their theories of the case early in the litigation and to adhere to those theories once they have been disclosed...." LG Electronics Inc. v. Q-Lity Computer Inc., 211 F.R.D. 360 (N.D. Cal. 2002) (quoting Atmel Corp. v. Information Storage Devices, Inc., No. C 95-1987 FMS, 1998 WL 775115, at *2 (N.D. Cal. 1998)). However, the expectation that a "patentee would have a precise sense of its infringement theory at the outset" is "unrealistic... [where] the patentee may not have been able to get access to the necessary information because it is hidden from view (for example, source code)." Peter S. Menell et al., Federal Judicial Center, Patent Case Management Judicial Guide 4-14 (2009). Thus, the good cause standard of Patent L.R. 3-6 "serves to balance the parties' rights to develop new information in discovery along with the need for certainty in legal theories at the start of the case." Apple, 2012 WL 5632618, at *2. "Courts typically grant leave to amend infringement contentions after a patentee has been given the opportunity to inspect relevant source code." Linex, 2013 WL 5955548, at *2.
Radware asserts that all its proposed amendments are based on information gleaned from Defendants' productions of source code. Good cause to amend exists because it has been diligent in reviewing the source code and in amending its infringement contentions based thereon. Moreover, Defendants will not be prejudiced because the AICs were served more than two months prior to claim construction briefing, and by citing to source code and eliminating several claims it previously contended were infringed, the AICs provide Defendants with greater specificity.
A. Compliance with Patent L.R. 3-1
As an initial matter, the Court will address the proper context for Defendants' arguments that Radware's AICs do not comply with Patent L.R. 3-1, which constitutes the majority of Defendants' oppositions. The Court agrees with Radware that this is not the appropriate vehicle for challenging the sufficiency of the AICs as a whole. Here, the Court is only concerned with (1) whether Radware acted diligently and (2) whether Defendants will be prejudiced by the amendments themselves as compared to the PICs. Thus, the sufficiency of the amendments is only relevant insofar as it affects the diligence and/or prejudice inquiries of the good cause standard.
F5 does not really fit its insufficiency argument into either prong and generally asserts that "Courts have found that one of the factors denying leave to amend is when the proposed amendments are still deficient." F5 Opp. at 6. However, in support it only cites a single federal claims court case, albeit one applying this district's Patent Local Rules, which noted that the amended claim chart was significantly deficient, but only after having already found that the plaintiff had not satisfied the good cause standard. See Canvs. Corp. v. United States, 107 Fed.Cl. 100, 109 (2012).
A10 asserts that the sufficiency of the Radware's AICs goes to the diligence prong of the good cause analysis. According to A10, "[i]t is axiomatic that a party cannot demonstrate diligence when it fails to satisfy the requirement of the local rules, " yet A10 cites no authority for support. A10 Opp. at 11.
The Court disagrees with the purported axiom and thinks that, if anything, the sufficiency of the amendments is more appropriately addressed under the prejudice prong of the analysis. While it is possible that amended infringement contentions could be so patently deficient as to evidence a lack of diligence (which is not the case here), it seems more likely that AICs might be insufficiently specific in violation of Patent L.R. 3-1 despite an earnest effort manifesting the requisite diligence. In the latter case, noncompliance may nevertheless cause the opposing party prejudice if, for example, the amendments would render otherwise sufficient infringement contentions noncompliant with the Patent Local Rules. Accordingly, to the extent that sufficiency of the AICs is relevant to this motion at all, it will be addressed under the prejudice prong of the analysis.
Defendants assert that Radware was not diligent in its review of source code after it was made available. A10 points to the fact that Radware took a two-week break from source code review from October 21 to November 1. F5 takes issue with Radware's eleven-day delay in starting its review. Moreover, the two and a half to three month period between source code production and service of the AICs also evidences a lack of diligence.
Radware explains that most of the delay was caused by constraints imposed by the stipulated protective order and was exacerbated by the Defendants' inactivity. Pursuant to the stipulated protective order, Radware could not bring its own computers into the review facility or take source code outside opposing counsel's office. Thus, in order to conduct its analysis and prepare the AICs, Radware had to request from Defendants hard copies of the portions of source code it wanted to analyze further. F5 took nearly three weeks to provide the requested printouts, while A10 was still producing theirs in piecemeal fashion as late as December 18.
Radware does not expressly address the gaps in its source code review at each facility, but presumably it focused on one defendant's source code at a time. In any case, continuous inspection of source code at both review facilities during the approximately six week period is not a prerequisite for diligence. In view of the massive amounts of code produced, the limitations imposed by the stipulated protective order, and Defendants' own failure to promptly produce hard copies of source code upon request, the Court thinks that the time spent at opposing counsel's offices reviewing the source code was reasonable, as was the additional time to prepare and serve its AICs based thereon. Accordingly, the Court finds that Radware exhibited the requisite diligence in amending its infringement contentions, at least insofar as the amendments were in fact based on the source code production.
1. Source Code Citations
It is undisputed that the citations to source code were necessarily dependent on the production of source code. Accordingly, Radware was diligent in amending its infringement contentions to include source code citations.
2. New Citations to Documents Previously Cited
Defendants assert that Radware cannot have been diligent in adding citations to documents it had in its possession by August 2013 at the latest; to have been diligent, Radware would have had to have included the citations in its PICs. Radware explains that any "supplemental analysis based on the same documents relied upon in the PICs are based on new insights gained by the review of previously unavailable source code."
The mere fact of prior possession does not automatically negate a finding of diligence, and the Court has no trouble believing that the subsequent production of source code could have, and in this case did, shed new light on information previously possessed. Accordingly, the Court is convinced that Radware acted at least minimally diligent in amending its infringement contentions to include citations to documents it previously possessed.
3. Doctrine of Equivalents
The parties raise essentially the same arguments for and against a finding of diligence with respect the newly asserted DOE theories. Defendants argue that because the DOE theories are not supported by citations to source code, they must not be based on the source code production. According to Radware, it was its "analysis of the source code in conjunction with [Defendants'] document production that informed Radware's DOE theories, which Radware now discloses in good faith." Again, the Court finds that Radware exhibited the requisite diligence in amending its infringement contentions to include some DOE theories.
Defendants' main argument applicable to the prejudice prong of the good cause analysis is that Radware's AICs are insufficient pursuant to Patent L.R. 3-1. Defendants primarily attack the sufficiency of the AICs as a whole, not just the proposed amendments. In doing so, Defendants are really challenging the sufficiency of the PICs as much as the AICs, which is not the Court's concern on this motion. Here, the Court's focus is on the amendments themselves. Thus, the relevant inquiry now is only whether Defendants will be prejudiced by the additions of citations to source code, citations to other documents, and the assertion of new DOE theories; not whether the amendments adequately cure any and all the purported deficiencies of the PICs. That being said, the Court does not think that Defendants are prejudiced by the amendments.
Defendants argue that they are prejudiced by Radware's citations to source code because Radware's purported pincites merely reference several pages of code, which is insufficiently specific per Patent L.R. 3-1. However, even assuming that Radware did not sufficiently pinpoint its citations, the Court is not convinced that Defendants are any worse off by having Radware cite to a few pages of code in the AICs as compared to a complete lack of identification of the source of infringement in the PICs. Thus, Defendants are not prejudiced by the citations to source code, at least not enough to outweigh Radware's showing of diligence.
Defendants do not contend that they will suffer prejudice by Radware's new citations to documents previously cited.
Defendants assert that they will be prejudiced by Radware's addition of boilerplate DOE theories. Although the Court thinks that an amendment to include assertions of new theories is an instance where noncompliance with the local rules may cause prejudice sufficient to negate good cause, that is not the case here. While Defendants argue that the DOE theories are insufficiently broad, they do not articulate how that translates to prejudice, and the Court does not think that the potential for prejudice here overcomes Radware's demonstration of diligence in seeking to add these theories based on Defendants' recent productions of source code.
The proposed amendments will not cause Defendants prejudice sufficient to outweigh Radware's showing of diligence. Accordingly, good cause exists, and Radware's motions for leave to amend its infringement contentions are GRANTED.
IT IS SO ORDERED.