Hynix Semiconductor Inc. v. Rambus Inc.

591 F. Supp. 2d 1038 (2006)

HYNIX SEMICONDUCTOR INC., Hynix Semiconductor America Inc., Hynix Semiconductor U.K. Ltd., and Hynix Semiconductor Deutschland GmbH, Plaintiffs,
v.
RAMBUS INC., Defendant.

Case No. C-00-20905 RMW.

United States District Court, N.D. California, San Jose Division.

January 5, 2006.

*1041 Kenneth L. Nissly, Susan van Keulen, Geoffrey H. Yost, Thelen Reid & Priest LLP, San Jose, CA, Theodore G. Brown, III, Townsend & Townsend & Crew LLP, Palo Alto, CA, Patrick Lynch, Kenneth R. O'Rourke, O'Melveny & Myers, Los Angeles, CA, for Plaintiff.

Gregory P. Stone, Kelly M. Klaus, Catherine Augustson, Munger, Tolles & Olson LLP, Los Angeles, CA, Peter I. Ostroff, Rollin A. Ransom, Michelle B. Goodman, Sidley Austin Brown & Wood LLP, Los Angeles, CA, for Defendant.

FINDINGS OF FACT AND CONCLSIONS OF LAW ON UNCLEAN HANDS DEFENSE

RONALD M. WHYTE, District Judge.

Hynix's unclean hands defense to Rambus's patent infringement claims was tried before the court on October 17-19 and October 24-November 1, 2005. The essential issues were (1) whether Rambus adopted a document retention plan in order *1042 to destroy documents in advance of a planned litigation campaign against DRAM manufacturers and (2) whether in light of any such conduct, the court should dismiss Rambus's patent claims against Hynix as a sanction for unclean hands. The court now issues its Findings of Fact and Conclusions of Law.

I. FINDINGS OF FACT

A. The Current Litigation

1. On August 29, 2000, plaintiffs Hynix Semiconductor America, Inc., Hynix Semiconductor, Inc., Hynix Semiconductor U.K., Ltd., and Hynix Semiconductor Deutschland, GmbH (collectively "Hynix")[1] filed a complaint, later amended, against Rambus Inc. ("Rambus") that in part sought a declaratory judgment of non-infringement, invalidity, and unenforceability of eleven Rambus patents. In February 2001, Rambus filed counterclaims asserting that Hynix infringed those eleven patents. Hynix subsequently answered the counterclaims and asserted various defenses. Rambus subsequently amended its counterclaims to add four additional patents.

2. The patents that have been asserted by Rambus in this case and their issue dates are set out in the following table:

-------------------------------------------------------------------
            Issue
Patent      Date      Title
-------------------------------------------------------------------
5,915,105   6/22/99   Integrated circuit I/O using a high
                      performance bus interface
-------------------------------------------------------------------
5,953,263   9/14/99   Synchronous memory device having
                      a programmable register and method
                      of controlling same
-------------------------------------------------------------------
5,954,804   9/21/99   Synchronous memory device having
                      an internal register
-------------------------------------------------------------------
5,995,443  11/30/99   Synchronous memory device
-------------------------------------------------------------------
6,032,214  2/29/00    Method of operating a synchronous
                      memory device having a variable
                      data output length
-------------------------------------------------------------------
6,032,215  2/29/00    Synchronous memory device
                      utilizing two external clocks
-------------------------------------------------------------------
6,034,918  3/7/00     Method of operating a memory
                      having a variable data output
                      length and a programmable
                      register
-------------------------------------------------------------------
6,035,365  3/7/00     Dual clocked synchronous memory
                      device having a delay time register
                      and method of operating same
-------------------------------------------------------------------
6,038,195  3/14/00    Synchronous memory device having
                      a delay time register and method of
                      operating same
-------------------------------------------------------------------
6,067,592  5/23/00    System having a synchronous
                      memory device
-------------------------------------------------------------------
6,101,152  8/8/00     Method of operating a synchronous
                      memory device
-------------------------------------------------------------------
6,324,120  11/27/01   Memory device having a variable
                      data output length
-------------------------------------------------------------------
6,378,020  4/23/02    System having double data transfer
                      rate and integrated circuit therefor
-------------------------------------------------------------------
6,426,916  7/30/02    Memory device having a variable
                      data output length and a programmable
                      register
-------------------------------------------------------------------
6,452,863  9/17/02    Method of operating a memory
                      device having a variable data input
                      length
-------------------------------------------------------------------

3. All of the patents-in-suit are continuation, continuation-in-part, or divisional applications based on a single parent application, serial number 07/510,898 ("`898 application").

4. In January 2005, Hynix moved for leave to add the defense of unclean hands to its pleadings. In an order dated March 7, 2005, this court granted that motion. In a separate order of the same date, the court ordered that Hynix's unclean hands allegations would be considered by the court in a separate, initial phase of the trial of the parties' respective claims.

B. The Farmwald/Horowitz Patent Applications

5. Rambus was founded in 1990 by two professors, Dr. Michael Farmwald and Dr. Mark Horowitz, who had been working together to address the increasing gap between microprocessor performance and *1043 dynamic random access memory ("DRAM") performance. Trial Transcript (hereinafter "Trial Tr.") 600:13-601:5; 1341:15-1343:2; 1540:12-19; HTX 005.001.[2]

6. From 1990 to the start of 2005, Geoffry Tate was the Chief Executive Officer of Rambus. Tate is presently the Chairman of the Board at Rambus. Trial Tr. 1226:9-16.

7. On April 18, 1990, Farmwald and Horowitz filed the '898 application. Trial Tr. at 364:11-365:19; 600:20-601:8; HTX 005.001.

8. The '898 application resulted in a number of continuation and divisional patent applications ("Farmwald/Horowitz family"). Rambus received its first issued United States patent resulting from the '898 application in September 1993. The patents that are at issue in this case resulted from this process. HTX 005.001

9. Rambus retained Blakely, Sokoloff, Taylor & Zafman ("BSTZ") as outside patent counsel from approximately 1981 through sometime in 2001 or 2002 to prosecute Rambus patent applications, including many applications from the Farmwald/Horowitz family. Trial Tr. 784:14-785:2. Lester Vincent, Scot Griffin, and Roland Cortes were patent prosecutors at BSTZ who worked on the Farmwald/Horowitz family of applications. Trial Tr. 784:14-785:2; 1592:22-1593:9; 1603:14-1604:11.

10. The first of the patents that Rambus has asserted against Hynix in this action issued on June 22, 1999. Rambus's Answer to Second Amended Complaint and Amended Counterclaim, filed 11/25/02, at 17-19.

C. Rambus's RDRAM Technology

11. Rambus does not manufacture its own products, rather, it licenses its intellectual property to DRAM manufacturers and collects royalties. Trial Tr. 1250:25-1251:2. As a company that generates revenue from its intellectual property alone, intellectual property protection is necessarily important to Rambus.

12. In the 1996-1999 time frame, Intel Corporation planned to use Rambus's RDRAM (Rambus Dynamic Random Access Memory) technology in its next generation of microprocessors. Trial Tr. 1237:20-1239:1. Because Rambus does not manufacture products, it relied upon DRAM manufacturers to license Rambus's intellectual property and produce RDRAM for use in Intel's products. Trial Tr. 1251:3-6.

13. Rambus referred to the RDRAM production by licensed DRAM manufacturers as the "Direct RDRAM ramp." Trial Tr. 1238:4-8; 1330:21-24. Direct RDRAM licensees were granted a narrow license to produce RDRAM. Trial Tr. 1289:1-16. These licenses generally did not permit licensees to utilize Rambus intellectual property for purposes other than producing RDRAM pursuant to Rambus's specifications. Other uses of Rambus's technology were referred to as "non-compatible" uses, because they were non-compatible with the RDRAM specifications. Trial Tr. 1356:22-1359:24.

D. Rambus's Participation in JDEC

14. Between 1992 to late 1995 or early 1996, Rambus was a member of the Joint Electron Device Engineering Council *1044 ("JEDEC"). Trial Tr. 786:21-795:8; 1148:11-12; 1161:12-20.

15. Richard Crisp, a program manager for Rambus, was one of Rambus's representatives to JEDEC and attended JDEC meetings on behalf of Rambus from 1992 to late 1995. Trial Tr. 1148:8-11. Billy Garrett was Rambus's other JEDEC representative. RTX 311; RTX 312.

16. Crisp and Garrett submitted trip reports following each meeting of JEDEC they attended. Crisp took a Macintosh laptop computer with him and took notes electronically. He later distributed his JEDEC trip reports to members of the Rambus executive team and others in the sales division.

17. Between 1992 and late 1995 or early 1996, Crisp, Tate, Tony Diepenbrock (inside patent counsel) and other Rambus executives and employees were informed that Rambus's participation in JEDEC might pose enforcement problems for some of its patents based on equitable estoppel for not disclosing to JEDEC Rambus's potential patent coverage of products (non-compatible with RDRAM) conforming to JEDEC standards. Trial Tr. 1156:4-1163:22; HTX 066; HTX 078; HTX 225. Rambus was also informed of this possibility by one of its outside patent attorneys, Lester Vincent. Trial Tr. 785:3-804:7; HTX 192. This concern was discussed within Rambus. Id.

E. MoSys License

18. In 1996, Rambus initiated patent licensing negotiations with Mosys. Mosys had been founded by two of Rambus's engineers and, according to Rambus, the MoSys product shared many characteristics with Rambus's products. Trial Tr. 1382:13-1384:17.

19. Patent negotiations, which Tate stated Rambus was poorly equipped to handle (Trial Tr. 1385:13-16), resulted in MoSys taking a license from Rambus. However, Rambus determined that since patent licensing negotiations were more complex than it had originally thought, it needed to hire someone familiar with negotiation of patent licenses to handle future licensing negotiations. Trial Tr. 1385:17-24.

F. Relationship with Hynix

20. Prior to 1998, Rambus and Hynix (then Hyundai) had a licensing agreement that included an "Other DRAM" provision, which apparently allowed Hynix to make non-compatible DRAMs using Rambus Interface Technology for a 2.5% royalty. Trial Tr. 1295:19-1305:11; HTX 004.004; HTX 087.

21. In July 1998, Rambus attempted to remove or amend the "Other DRAM" provision. Id. The reasons apparently were that Rambus wanted Hyundai to increase its marketing effort and productize the RDRAM device and to be able to claim infringement by Hyundai if it continued to work on SL DRAM (Synchronous Link DRAM). Id.; Trial Tr. 1038:2-1039:22.

22. Thereafter, Hynix merged with LG Semiconductor ("LGS"). At the time of the merger, both companies had licenses with Rambus. Id. Hynix attempted to affirm the 1995 Hynix-Rambus license, asking Rambus to agree to use the royalty rate specified in the Hynix-Rambus license. Id.

23. Rambus apparently decided to treat the LGS license as governing the relationship between Rambus and the merged company. That agreement apparently did not have an "other DRAM" provision." Trial Tr. 1038:2-1039:22.

G. Rambus Formulates its Licensing and Litigation Strategy

1. Rambus Hires Joel Karp

24. Rambus hired Joel Karp in October 1997 to assess its patent portfolio, determine *1045 if chips infringed the patent portfolio, develop licensing strategies for infringing chips, and negotiate with companies that built and sold such chips. Trial Tr. 356:22-357:23; RTX 080; HTX 091. In advance of the commencement of Karp's employment, Geoff Tate, Rambus's CEO, sent an e-mail to Rambus executives describing Karp's role as "to prepare and then to negotiate to license our patents for infringing drams (and potentially other infringing ic's)." HTX 091; RTX 080.

25. Before his employment at Rambus, Karp was employed by Samsung from September 1990 through July 1997. When he left Samsung's employment, he was a senior vice president. During his employment with Samsung, Karp attended JEDEC meetings on Samsung's behalf, describing his role as "Samsung's mouthpiece." Karp met Richard Crisp, Rambus's JEDEC representative, at JEDEC meetings. Trial Tr. 136:16-25.

26. Karp had learned through his experience that the DRAM industry was very litigious. Trial Tr. 138:23-134:3.

27. While at Samsung, Karp participated in licensing and litigation activities on behalf of Samsung. Trial Tr. 138:2-22. In one action against Texas Instruments ("TI"), Karp submitted a declaration asserting that TI was subject to equitable estoppel because it was contrary to industry practice for an intellectual property owner to remain silent during the standard setting practice if its intellectual property covered the standard being considered. Trial Tr. 150:12-151:7.

28. On January 7, 1998, Tate met one-on-one with Karp. Karp was instructed to prepare a plan for licensing infringing DRAMs for presentation to the Rambus Board of Directors in early March. That strategy was to include a litigation strategy. Trial Tr. 170:2-171:12; HTX 013.020; see also HTX 395.02 (Cooley Godward attorney Peter Leal's notes from a January 13, 1998 meeting with Tate and Karp stating "Want litigation strategy by March board meeting. Six weeks from now.").

29. According to Tate's top level goals for 1998, Rambus's IT goals included positioning Rambus intellectual property for the future. Under the heading "Position Rambus for the Future Including IP," Tate's top level goals included the following:

Develop and enforce IP

A. Get access time register patent issued that reads on existing SDRAM
B. Broad patents in place for Direct Rambus, next generation signaling; and chip-to-chip interconnect
C. Get all infringers to license our IP with royalties > RDRAM (if it is a broad license) OR sue.

HTX 094.

2. Rambus Meets With Cooley Godward Attorneys

30. In late 1997, Karp called Diane Savage, an attorney at the law firm of Cooley Godward ("Cooley") with whom he had worked before coming to Rambus, seeking a recommendation for someone to help set up a licensing program. Trial Tr. 393:22-394:10. Attorney Savage, who was a partner in Cooley's technology transactions group, introduced Karp to her colleagues John Girvin, Dan Johnson, and Peter Leal. Trial Tr. 394:17-22; 585:16-19.

31. In January and February 1998, Karp began to meet with the Cooley lawyers to discuss issues relating to patent licensing. HTX 007; HTX 376; HTX 395. At the January 13, 1998, January 15, 1998, and February 12, 1998 meetings between Cooley attorneys and Rambus executives, Tate and Karp, the discussion about a licensing strategy included formulating a litigation strategy as a part of a licensing strategy. HTX 376, HTX 395, HTX 403.

*1046 32. On January 13, 1998, Karp and Tate met with Cooley attorney Leal, an attorney specializing in licensing matters. On January 15, 1998, Karp and Leal met again. HTX 376, HTX 395.

33. At the January 13, 1998 meeting, the parties discussed the concept "[n]o negotiation w/out full strategy and prep." Rambus wanted to "go in and quickly proceed to either a license or litigation." Further, Rambus was "looking for a royalty rate that tells [the DRAM industry] it costs to infringe." Rambus wanted to "[t]ry win-win first; do not prejudice g-f for litigation." HTX 395.

34. At the January 15, 1998 meeting, Karp and Leal discussed a proposed sequence for negotiating meetings with potential infringers, including roles Rambus executives and Cooley attorneys might play in negotiations and what information would be presented at each meeting, labelled in Leal's notes as the "Middle Ground, delaying meeting" and the "Pound Sand" meeting. Rambus at this time was "very, very sensitive" to costs. HTX 376.

35. On February 12, 1998, Karp met with three Cooley attorneys, Johnson, Girvin, and Leal. Leal, a licensing attorney, reported to Girvin, head of the Information Technology Patent group. Trial Tr. 586:9-16. Girvin, in turn, reported to Johnson, head of Cooley's litigation group.

36. The purpose of the February 12, 1998 meeting was to develop the licensing strategy Tate had requested at the one-onone meeting between him and Karp. HTX 097; HTX 403. The licensing strategy envisioned optimizing Rambus's notice to potential infringers, a negotiation strategy and a litigation strategy. HTX 403.

37. As of the February 12, 1998 meeting, the Cooley attorneys were aware of Karp's draft licensing term sheet, which specified that Rambus would charge a 5% running royalty for a license to make noncompatible DRAMs. RTX 088. Karp acknowledged that these royalty rates "will probably push us into litigation quickly." HTX 097.

38. At the February 12, 1998 meeting, Johnson expressed concern about Rambus having no document control system in place.

Q. Okay. The next bullet point says "Make ourselves battle ready." Do you see that, that's the first part of it?
A. Correct.
Q. Do you recall somebody saying words to that effect at this meeting?
A. That sounds like something I would do.
Q. And what did you mean by that phrase?
A. Very simply, Rambus was essentially an old start-up, as far as I was concerned. It had been around probably eight, nine years. They had, as best I could tell, no central document control system in place . . . and if you want to have a licensing program, if you end up filing lawsuits against anybody, putting in place a system that gets your documents organized.
* * *
Q. And the last sentence in this bullet point says "Need company policy on document retention policy." Do you see that?
A. Correct.
Q. Is this something you said at this meeting?
A. Absolutely.
Q. . . . What led you to say that to Rambus?
A. Typical start-up. As I said, they didn't have a policy, as best I could tell. I can't recall if somebody mentioned it to me at this point in time, but I'd just litigated a case where a company was, either had to spend, I can't remember if it was $60,000 or $100,000 trying to retrieve *1047 data, and the President was screaming. And they'd gone back through and recycled and they didn't have a document retention policy. They got accused of spoliation. So you get yourself a document retention policy.

Trial Tr. 1675:21-1678:20.

39. Johnson advised Rambus to gather critical company documents to start putting together an electronic database and to adopt a company policy on document retention. HTX 097. Johnson testified that he recommended adopting a document retention policy for three principal reasons. First, a document retention policy would reduce the expense of retrieving electronic data stored on obsolete or corrupted backup media. Trial Tr. 1676:24-1677:10. Second, a document retention policy would reduce search costs in the event that Rambus was someday required to respond to subpoenas or document requests that might possibly be issued in connection with future lawsuits or investigations. Trial Tr. 1677:11-14. Third, the absence of a company-wide policy for the retention and destruction of documents might be cited by a future litigant as evidence of spoliation. Trial Tr. 1678:10-20.

40. Johnson testified that the advice he gave Rambus regarding document retention was commonplace and that he probably gave similar advice to at least eight to ten start-up companies in the Silicon Valley. Trial Tr. 1678:21-1679:9.

41. At this February 12, 1998 meeting, Johnson also advised Rambus to instruct its patent prosecution attorneys to clean up their files for issued patents to ensure that the file was the same as the official file, a recommendation that Johnson characterized as "standard advice." Trial Tr. 409:18-410:1; 1679:10-1680:3; HTX 097.

42. On February 23, 1998, Cooley presented its "Proposed Strategy for Rambus." HTX 098. The document briefly outlines a licensing strategy, starting with criteria for selecting initial targets for negotiation. The proposed strategy progresses to a discussion of a tiered litigation strategy. One of the basic assumptions was that Rambus would not initiate action until a competing product enters the market, at which time Rambus would conduct reverse engineering and determine what action to take next. HTX 098.002.

43. Cooley's proposed strategy offered two options: first, pursuing a breach of contract remedy against existing licensees; second, initiating a patent infringement suit against an unlicensed competitor; third, bringing an action against SLDRAM. With respect to possible litigation, the proposed strategy memo states:

To implement the above strategy, Rambus has authorized outside counsel to begin organizing documents and preparing a discovery data base, so that if and when Rambus elects to proceed with litigation, it will not unduly disrupt the company's activities. More importantly, with proper planning, Rambus may be able to obtain an advantage over its competitors by choosing a court such as the eastern district of Virginia, "the rocket docket" or the ITC. Because these courts proceed at an accelerated schedule, early preparation will benefit Rambus.

HTX 098.002-003.

44. The Cooley document does not specifically mention establishing a document retention policy at Rambus, but does state that "Rambus has authorized outside counsel to begin organizing documents and preparing a discovery database, so that if and when Rambus elects to proceed with litigation, it will not unduly disrupt the company's activities." HTX 098.002. The Cooley Proposed Strategy for Rambus ends stating "it bears emphasis that each of the above scenarios is dependent on the facts that exist at the time the decision to litigate *1048 is made. Factors unknown at this time may result in a change of strategy." HTX 098.003. Karp added in handwriting "document retention policy" and "patent attorney files" after the text of the proposed strategy.

45. On February 25, 1998, Karp had a regular one-on-one meeting with Tate at which they reviewed the developing licensing and litigation strategy. HTX 013.034.

3. Karp Presents a Licensing and Litigation Strategy to the Board

46. On March 4, 1998, Karp presented the licensing strategy to the Rambus Board of Directors. HTX 031. The strategy included demanding the 5% running royalty rate and other financial terms that Karp had reviewed with Cooley at the February 12, 1998 meeting. HTX 006.001.

47. The minutes following the March 4, 1998 board meeting memorialize that the meeting included an update by Karp of "the Company's strategic licensing and litigation strategy." HTX 031.002.

48. In the presentation, Karp recommended that "[i]f licensing discussions do not result in resolution, tiered litigation strategy kicks in." HTX 006.003. He suggested that the first targets for licensing non-compatible uses should be "present licensees which currently have released non-compatible product," and that the second group of targets should be "present licensees which are currently well along with alternate development." HTX 006.006.

49. The March 4, 1998 presentation included a timetable for executing the proposed licensing and litigation strategy. The first milestone was the delivery of initial samples of Direct RDRAM ("RDRAM"). This was to ensure that DRAM manufacturers would be locked into the RDRAM ramp. HTX 006.007. Thereafter, the strategy included procuring customer sample ("CS") quality parts of potentially infringing devices, reverse engineering the products and creating claim charts, notifying the potential infringer, conducting two meetings, and if the meetings did not result in agreement on a license, commencing legal action. The timetable projected 4-6 months from procuring "CS quality parts" to the commencement of litigation, HTX 006.007.

50. At the time of the March 1998 board meeting, Rambus's anticipation was that the D-RDRAM would not be ready for high volume production until late 1998. HTX 094 (Item 2: "Direct Rambus 1.0 memory system implemented on spec, on cost, on schedule for production ramp late 98"); Trial Tr. 1250:1-11.

4. Rambus Revises its Strategy

51. In late 1998, Rambus revised its approach toward licensing non-compatible technology to its existing licensees. Rambus initially planned to begin its licensing strategy only at the time the DRAM manufacturers were locked in to RDRAM production. By October 1998, the projected time frame for this was early 2000. Karp in an October 1998 strategy update projected that Rambus might be able to demonstrate that Mosel and Nanya had SDRAM products that directly infringed a pending access time register patent by the first quarter of 1999, thus enabling Rambus to potentially state contributory infringement and inducement claims against companies like Acer, SIS, and VIA for SDRAM and DDR. However, the presentation recommended not even initiating licensing negotiations.

▄ DO NOT ROCK THE DIRECT BOAT.
▄ We should not assert patents against Direct partners until ramp reaches a point of no return (TBD)
▄ Probably not until Q1'00
* * *
*1049 ▄ However, the Big Question Is— WHAT'S THE RUSH?
▄ What is the compelling business reason? I can't think of any.
▄ Keeping the Maytag repairman busy is not a valid reason
▄ IMHO, risks of damaging establishment of dominant standard outweigh potential return
▄ Lets not snatch defeat from the jaws of victory

HTX 128.003; Trial Tr. 1319:17-1321:5; 282:20-284:2. This "point of no return" was projected to be in the first quarter of 2000. Id.

52. Karp's October presentation recommended that Rambus's "Top priority Should Be Strengthening of Portfolio" by covering SDRAM, DDR, SLDRAM, any and all forms of synchronous memory (static and dynamic) by aggressively prosecuting claims that provide such coverage. HTX 128.005. The projected date by which Rambus's "strategic portfolio" would be "ready for presentation to the industry" was calendar year 2000. HTX 128.006.

53. In December 1998 or January 1999, Karp drafted and distributed to Rambus executives a "Nuclear Winter Scenario" memorandum relating to Rambus's Patent Enforcement strategy for 1999. That memo set forth a hypothetical scenario in which Intel suddenly opted to move away from Rambus's technology, instead implemented something else, for example DDR or SLDRAM. Karp set forth a strategy for convincing Intel "that without access to Rambus' IP, it will be difficult and costly to continue selling its current processor based products and its new, more advanced products because the memory needed for these Intel products require use of Rambus' IP." HTX 004.2.

54. The Nuclear Winter Scenario memorandum proposed that Rambus could establish its intellectual property position by showing by clear and convincing evidence that three of Rambus's then-issued patents covered alternate competing devices: the '327 patent covered DDR (dual edged clocking); the '481 covered DDR (phase locked loop circuitry); and the '580 covered DDR and PC100 (access time register). HTX 004.002.

55. In a strategy update presentation titled "IP Strategy 9/24/99," Rambus recognized that it might be losing Intel's business ("Intel Has Already Started To Let Go") and that, absent that relationship, the DRAM industry did not have respect for Rambus's intellectual property. HTX 244. The presentation stated that Rambus must "earn that respect by substantiating our claims that cover pioneering technology" and that "[p]atent claims are substantiated either by

▄ signing a lucrative license deal with (an) industry powerhouse(s)
▄ winning in court"

HTX 244.004.

H. Rambus's Formulation of its Document Retention Policy

56. At a one-on-one meeting with Tate on February 25, 1998, Karp reported that he was already "[w]orking with J. Lau on a document retention policy, discovery database, email, etc." HTX 013.034. Tate did not recall this meeting with Karp. Trial Tr. 1275:8-12. However, it was Tate's practice to meet with his executive staff one-on-one every week to discuss goals and progress toward goals. Tate's inability to recall discussing document retention and litigation strategy as part of licensing strategy suggests that he either was focused on issues other than document retention or wants to avoid being involved in the current phase of the litigation.

57. On March 16, 1998, Allen Roberts, Vice President of Engineering, advised Joseph Lau, Engineering Department Head, that Rambus should begin recycling its e-mail *1050 backups so that the backups had a "shorter shelf life." HTX 100. Roberts suggested three months. Id. Roberts understood a concern that Rambus's e-mails were discoverable. Trial Tr. 1575:17-1577:4

As I recall, the general concept was... that we were generating gigabytes of information on surely a weekly basis, but maybe even on a daily basis that went onto tapes.
And that those tapes were being kept kind of continuously, and that if there was any case where somebody said we want to see all this information, it would be a huge expense to pull all that information back and pay people to review it all.

Trial Tr. 1577:7-18.

58. In early 1998, while Rambus was separately consulting with outside counsel regarding the adoption of a document retention policy, Rambus retained Kroll Associates, a specialist in the field of information and computer security, to do a security audit of Rambus. Rambus contacted Kroll because it had security concerns after a hacking attempt on Rambus's computers. Trial Tr. 369:18-370:20. In an April 24, 1998 presentation at the conclusion of the security audit, Kroll advised that Rambus adopt a variety of measures to better protect its confidential information and the confidential information of its business partners. September 20, 2005 Deposition Transcript of Alan Brill at 49:15-50:1; 50:13-18 (played at trial on October 28, 2005, Trial Tr. 1515:21-23, 1527:8-9). Among the specific recommendations from Kroll was the confirmation that Rambus should work with outside counsel to develop a document retention policy. Trial Tr. 372:25-373:23; RTX 160 at 26.

59. On March 19, 1998, after an inquiry about document retention policy by Karp, Attorney Savage forwarded to Karp a seven page standard document retention policy template that had been developed by Cooley. Trial Tr. 579:19-580:19, 582:19-24; RTX 091. Attorney Savage was unaware of any litigation strategy at the time she forwarded the template. Trial Tr. 583:15-17. In her forwarding memorandum to Karp, she recommended that he consult Cooley attorney David Lisi if he had any specific litigation oriented issues. Trial Tr. 579:19-584:2; RTX 091 at 1.

60. The memorandum sent to Rambus by Cooley was a generic template for a document retention program drafted by the law firm for its clients. Trial Tr. 580:15-19. Karp did not consult Attorney Savage or Lisi regarding tailoring the template for Rambus.

61. On March 27, 1998, after he had left Cooley to join Fenwick & West, Attorney Johnson had a lunch meeting with Karp. Following that lunch, Johnson wrote Karp saying "I am excited about the possibility of working with you." Attorney Johnson also enclosed "a standard set of document requests," saying "[t]his should give you some idea of the type of information requested in patent cases." HTX 368; Trial Tr. 1713:15-1716:9.

62. The Cooley template retention policy stated that "`[t]hat Policy should inform employees email is subject to review by the Company and is not `private.'" RTX 091 at R401092. The Cooley template also recommended that the company should "permanently remove emails from system server on a periodic basis" and if back-up tapes of e-mail are kept, they "should be destroyed on a periodic basis as well." Id. It also stated that "the Company and individual employees should be discouraged from archiving email" and that "E-mail that needs to be saved should be either: (a) printed in hard copy and kept in the appropriate file; or (b) downloaded to a computer file and kept electronically or on *1051 disk as a separate file." Id. The memorandum also recommended that Rambus "regularly discard outdated electronic files and discard all draft documents once a document is finalized." Id. Brill of Kroll Associates similarly recommended that Rambus keep e-mail only as long as required and that backup tapes should not be kept indefinitely. RTX 160 at 26, 47; September 20, 2005 Deposition Transcript of Alan Brill at 55:19-57:14 (played at trial on October 28, 2005, Trial Tr. 1515:21-23, 15:178-9).

63. On May 14, 1998, Karp met with a group of representatives of the four operating divisions to decide on a policy for saving back-up tapes. He then sent a memorandum to the Rambus board, engineering managers, and others at Rambus stating: "Effective immediately, the policy is that full system backup tapes will be saved for 3 months only. Therefore, you can no longer depend on the full back ups for archival purposes. Any valuable data, engineering or otherwise, must be archived separately." RTX 104. Karp had consulted with Attorney Johnson about Rambus's policy to retain backup tapes for 3 months. Trial Tr. 528:21-530:4; 1691:23-1692:6; RTX 104. Johnson recommended against adopting a proposal to create a full system backup every two years, citing the problems and potential expense that could arise from stale data, operating system changes, and corrupt data. Trial Tr. 1692:7-17.

64. Karp drafted Rambus's Document Retention Policy shortly before July 22, 1998, based upon the memorandum forwarded by Attorney Savage, and incorporating some of its language verbatim. Trial Tr. 525:7-17; HTX 023. The Rambus policy provided with respect to "Electronic Mail and Documents" that "Rambus maintains complete system tape back-ups for a period of 3 months. Employees should not utilize email as a place to save documents beyond 3 months. Email that is required to be saved more than 3 months can be kept either in paper or a separate file on your hard drive." HTX 023.

65. The Rambus Policy stated that:

documents, notebooks, computer files, etc., relating to patent disclosures and proof of invention dates are of great value to Rambus and should be kept permanently. Engineering personnel should not depend on the electronic system back-up tapes to archive their work, since these tapes are only kept for 3 months. They should create their own archive copies of, for example, tape out and major project milestone databases, which can then be kept indefinitely in Rambus' offsite, secure storage facility.

HTX 023.001-002.

66. The Rambus policy pronounced that "Final, execution copies of all contracts entered into by Rambus are kept at least 5 years after the expiration of the agreement.... All drafts ... and any materials used during negotiations that are not part of the final binding contract ... should be destroyed or systematically discarded." HTX 023.002

67. The terms of Rambus's two page Document Retention Policy were based on the template provided by Attorney Savage, referred only to categories of documents, and were content neutral within those categories. HTX 023; RTX 091. The Policy contained no directive to discard documents relating to specific companies or to certain subjects. The evidence does not suggest that Karp used the document request sample provided by Attorney Johnson to tailor Rambus's Document Retention Policy language to target particular categories of harmful document.

68. Hynix is not challenging the propriety of any of the written provisions of the Rambus document retention policy itself *1052 nor arguing that any of the provisions is inconsistent with industry custom or practice. Trial Tr. 7:15-8:3.

I. Presentation of Rambus's Document Retention Policy

69. After Karp had drafted the Policy, he created slides for use in a presentation introducing the document retention policy to Rambus's managers. Karp sent his two-page document retention policy to Attorney Johnson for his review. Trial Tr. 521:20-522:10; 1692:18-1693:2; 1693:13-25.

70. Only July 22, 1998, Kambus distributed its two-page document retention policy to Rambus employees. HTX 023.

71. On July 22, 1998, Attorney Johnson and Karp gave a presentation to Rambus's managers about the need for such a policy and the legal requirements involving such policies. RTX 130.

72. Karp's portion of the presentation consisted of a summary of the types of documents Rambus wanted to keep and for how long. Attorney Johnson reviewed and commented on drafts of Karp's slides. Trial Tr. 535:4-11; 1699:11-1701:4; RTX 112; RTX 115; RTX 123. Karp in presenting his slides to Rambus managers and employees gave explicit instructions to delete e-mails. Trial Tr. 261:4-265:1.

73. Attorney Johnson's portion of the presentation was litigation-oriented. Slide 3 lists what kind of records are discoverable. Slide 4 cautions specifically about email and electronic documents. Slide 5 recounts what Johnson referred to as "horror stories" of cases where deleted e-mails had been used to prove age discrimination and sexual harassment. Slides 6 and 7 warn that the rules for document retention change once litigation begins. RTX130; HTX111.004-008. Attorney Johnson testified that "[a]nd the act of deleting, when you are actively involved in litigation, that can give rise to a claim, and I was trying to explain to the client that this was serious business." Trial Tr. 1699:7-10. Johnson also advised that "[a] formal document retention policy will likely shield a company from any negative inferences ... due to destruction of documents, unless the policy was instituted in bad faith or exercised in order to limit damaging evidence available to potential plaintiffs. RTX 130 at R124523. He also warned Rambus managers that destroying relevant documents once litigation started would be improper. RTX 130 at XXXXXX-XX, XXXXXX-XX; Trial Tr. 1722:2-11.

74. The slides that Karp prepared for Attorney Johnson's review repeatedly directed Rambus employees to "look for things to keep." HTX 112. Johnson told Karp that such a directive would result in the retention of more documents than Rambus employees were otherwise required to keep. Trial Tr. 1700:13-1701:1. Karp nevertheless chose to keep the language in his slides. Trial Tr. 1701:2-4.

75. Following the July 22, 1998 presentation by Johnson, Johnson no longer assisted Rambus in its licensing, litigation, or patent prosecution strategy. Karp merely asked Johnson "truly random" questions from time-to-time about general litigation related matters; such as arbitration and different venues for litigation. Trial Tr. 1719:20-1721:17; RTX 348; RTX 350; RTX 354.

76. After presenting the document retention policy jointly with Attorney Johnson to the Rambus managers, Karp made presentations to various groups of Rambus employees using a different set of slides, which were based on the written policy. Trial Tr. 262:19-263:5; HTX 112.

77. All versions of the presentation utilized a slide discussing e-mail entitled "Email—Throw it Away." The slide emphasized that e-mail was discoverable in litigation and that elimination of e-mail is *1053 an integral part of document control. It also advised that messages that must be saved should be saved to a separate file or printed and retained. HTX 111.010.

J. Specific Events Involving the Destruction of Documents

1. Shred Days

78. On September 3 and 4, 1998, Rambus employees participated in a companywide "shred day" ("Shred Day 1998"). Rambus employed an outside company, Sure Shred, to provide on-site document shredding services. See HTX 125; Trial Tr. 1614:16-1615:11. Employees were instructed to follow the Document Retention Policy guidelines to determine what to keep and what to throw away. Trial Tr. 272:14-273:14; HTX 122. Employees were given burlap sacks from Sure Shred for material that needed shredding. The burlap sacks were then taken to a shredding truck in the parking lot of the company. Id.

79. Rambus destroyed approximately 185 bags and 60 boxes of material on Shred Day 1998. HTX 125. The amount of documents destroyed would have totaled, on the high side, 430 banker boxes of material. Trial Tr. 1615:12-1616:18.

80. A year later, on August 26, 1999, Rambus had another housekeeping event, or "shred day" ("Shred Day 1999"). The bulk of the documents destroyed on the two shred days was "mountains of printouts of circuit design that engineers had." Trial Tr. 337:10-338:3.

81. The use of burlap bags and shredders, the volume of material shredded, and the social events at the end of the "shred days" were unexceptional. Prior to Shred Day 1998, each employee had a box in his or her office area for confidential documents that needed to be shredded, and a truck came once a week to pick up the documents for shredding. Trial Tr. 1581:13-1582:10. There was nothing unusual about the volume of material that Rambus shredded. In particular, David Rhoads of Sure Shred suggested that Rambus's disposal practices were no different from the disposal habits of other similarly situated companies. Trial Tr. 1612:5-1623:15. Rhoads testified that his company frequently provided pick up services "where it's time to purge records on an annual basis, or [a company has] just begun a document destruction program so they call us out to do a big purge." Trial Tr. 1612:24-1613:2. He observed that on these annual "purge days," companies often have an "event" and serve food as part of that "event" as Rambus did. Trial Tr. 1613:18-21. A contemporaneous news article offered by Hynix also demonstrates the commonplace nature of both "shred days" and of social events associated with them. HTX 130 (a November 11, 1998 article from The New York Times that refers in part to an "event" at Amazon where employees were asked to "purge" emails that were no longer needed for business or legal purposes). See also David O. Stephens and Roderick C. Wallace, Electronic Records Retention, An Introduction 19, Chapter 7—Implementation Issues, at 48 (Arma International 1997) ("Setting aside one day a year for office workers to clean out files, destroy old records, and transfer records to the records center is popular in some organizations. Management allows people to `dress down,' and provides a free lunch on purge day. Sometimes awards are given to those persons and departments that recycle the most paper.").

2. Electronic Documents

82. Until approximately 1996, business work was done on Macintosh computers. Engineering work, by contrast, was typically done on UNIX workstations. Trial Tr. 1439:4-6, 1440:2-11. Rambus maintained backup tapes of some of the data *1054 created on the Macintosh computers. In May 1998, Karp confirmed that backup tapes of the company's Macintosh computers were gone. Trial Tr. 208:19-218:14; HTX 013.097.

83. On or before July 21, 1998, Rambus sent 1269 computer tapes out to be degaussed (magnetically scrambled to prevent data recovery). Trial Tr. 221:22-223:4, 1105:25-1106:25; HTX 107, HTX 157, HTX 029. Prior to sending the tapes to be degaussed, Rambus separated its backup tapes into two sets: (1) project tapes containing only technical project data and (2) other tapes. The "other tapes" were sent for degaussing without further review of the contents. Trial Tr. 1105:17-1106:25.

3. Prosecution Files

84. Attorney Johnson advised Karp at the February 12, 1998 meeting that Rambus should clean out its patent prosecution files so that the files mirrored the Patent Office ("PTO") file. Hynix does not allege that there is anything improper in the general practice of cleaning patent files upon issuance. Trial Tr. 837:12-22.

85. Attorney Vincent of BSTZ prosecuted patent applications on behalf of Rambus claiming priority to Rambus's original '898 application, as did other BSTZ attorneys including Roland Cortes and Scot Griffin. Trial Tr. 784:1-785:2; 1592:22-1593:9; 1603:19-1604:11.

86. In April 1999, Karp instructed Vincent to clean BSTZ's patent prosecution files for patents which had issued. Trial Tr. 806:5-807:5; 812:3-14; HTX 135.004. Vincent was not told that any litigation was planned against manufacturers of JDEC standard SDRAM or other non-conforming DRAM. Trial Tr. 812:15-21. Karp did not provide details as to what documents should be discarded from the files; rather, Vincent was guided by his "general professional knowledge" regarding document retention for patent files. Trial Tr. 904:11-24.

87. Vincent was not instructed to, and did not, clean any of BSTZ's "general" Rambus files. Trial Tr. 908:25-909:7. The general files contained, among other papers, Vincent's notes and other documents relating to his advice to Rambus concerning the JEDEC disclosure policy and equitable estoppel. Trial Tr. 908:20-24 (See, e.g., HTX 192.001 (March 27, 1992 notes by Lester Vincent referring to JDEC and equitable estoppel)). These documents were retained and produced. Trial Tr. 909:10-910:6; 911:9-25; RTX 179 (January 31, 2000 cover letter from Vincent to Neil Steinberg transmitting copies of various BSTZ general files).

88. Some drafts and notes from BSTZ's patent prosecution files which were copies or otherwise not important were discarded as a matter of common practice of BSTZ's attorneys before Karp made his request to Vincent. Trial Tr. 905:12-906:8; 1607:19-609:5. Attorneys Griffin and Cortes understood that it was the firm policy at BSTZ to clean patent files upon issuance with respect to all patent files. Trial Tr. 1598:2-1600:19; 1606:18-1607:10. Griffin left BSTZ in April 1996, approximately three years before Karp's request to Vincent, and Cortes left BSTZ in April 1999, at approximately the time of Karp's request. Trial Tr. 1592:11-14; 1603:14-18. Cortes and Griffin each prosecuted to issuance Rambus patent applications claiming priority to the original Rambus patent application, including at least three, namely, U.S. Patent Nos. 5,513,327, 5,657,481, and 5,841,580. These files presumably would have been purged by them after issuance of the patents. Trial Tr. 1593:10-1598:1; 1604:12-1606:17; HTX 005A (indicating that the '327, '481, and '580 patents correspond to *1055 BSTZ file numbers P001C2, P007DC and P043D2, respectively).

89. BSTZ's patent prosecution files were organized into three parts: left, center, and right. The left side had correspondence and notes, the center typically had prior art, conception, and reduction to practice documents and drafts, and the right side had communications with the PTO. Trial Tr. 895:3-896:11. When cleaning Rambus patent prosecution files for issued patents, Vincent retained, among other things, conception and reduction to practice documents and communications with the PTO, which constituted the right side of the file, and prior art. Id. The portions of the files from which documents were discarded came from the left and center portions. Id.

90. On February 1, 2000, Vincent provided copies to Rambus of the left and center parts of the files of all but one of BSTZ's patent prosecution files for applications claiming priority to the '898 application. Trial Tr. 912:1-23; RTX 181 (February 1, 2000 cover letter from Lester Vincent to Neil Steinberg enclosing copies of files). The remaining file had been omitted due to an oversight and was provided later. Trial Tr. 913:3-17.

91. Vincent did not clean any Rambus patent files between July 28, 1999 and June 23, 2000. HTX 327. Vincent's renewed cleaning of patent files on or after June 23, 2000 apparently had no impact on the documents produced in this litigation. As noted above, copies of any documents from the files of patents claiming priority to the '898 application that Vincent may have discarded on or after June 23, 2000 had already been provided to Rambus. Trial Tr. 914:3-12.

92. No material documents have been rendered unavailable to Hynix from the files of any of the patents-in-suit as a result of Karp's request to Vincent. Of the 15 patents that have been asserted by Rambus in this litigation, Hynix has alleged that the files of three of them—U.S. Patent Nos. 5,915,105, 5,954,804, and 6,101,152—were cleaned by Vincent. HTX005A. These patents correspond to internal BSTZ file numbers P001C5, P010DCD, and P043D2C2D1 and were among the files reviewed by Vincent on June 23, 2000, after copies of any documents that might have been discarded from the file had been provided to Rambus. HTX 327. Also, no evidence suggests that any material evidence was discarded from any prosecution files. Prior art was retained.

K. Rambus Implements its Licensing Strategy

93. On July 10, 1998, Tate sent an internal e-mail to Roberts and Karp regarding the license agreement between Hyundai and Rambus. As part of a discussion of revised license terms, Tate suggested that Hyundai would be "a great company to start Joel's plan with in q1/99 potentially." HTX 087.

94. Karp's Strategy Update 10/98 presentation suggested that there was no rush to begin seeking non-compatible licenses until "ramp reaches a point of no return." HTX 128.004. This "point of no return" was projected to be in the first quarter of 2000. Id.

95. In the summer of 1998, Attorney Neil Steinberg first did a small assignment for Rambus. He began working again for Rambus in mid-August 1998 and he continued until April 1999 at which time he became Rambus's General Patent Counsel. Trial Tr. 1443:2-1444:20. Starting in October 1998 Steinberg's work included prosecuting continuation applications for the Farmwald/Horowitz family of patents. Trial Tr. 1463:7-1464:13.

96. Karp's intellectual property goals for the third quarter of 1999 set forth in a *1056 document dated July 1, 1999 included presenting his licensing strategy to Rambus's board for approval in the fourth quarter. These goals also included preparing licensing positions against three manufacturers and a litigation position against one of those three licensing targets. HTX 139.001.

97. Karp and Steinberg developed a formula for selecting licensing targets. The formula was set forth in a presentation dated October 14, 1999. HTX 151. The formula considered both business and legal factors. Among the factors weighed were "confirmation of Rambus IP," "experience in battle," "exposure to Rambus IP," and "litigation story." HTX 151.002. Hynix was identified as the number three target after Hitachi and Samsung. HTX 151.006.

L. Rambus Executes its Litigation Strategy

98. At the beginning of each year, Rambus would go through a budgeting process to allocate where money would be spent. Tate testified that through 1999, expenditures over $5,000 to $10,000 would be in the budget. Trial Tr. 1570:1-1571:13. In 1998, Attorney Johnson told Rambus that the cost of pursuing litigation against a DRAM manufacturer would likely cost between $1.5 and $3 million. Trial Tr. 518:13-519:3; HTX 004.006; RTX 354 at R125823. Rambus did not budget any money in 1998 or 1999 for litigation against a DRAM manufacturer. Trial Tr. 376:23-378:21, 383:15-18, 383:23-384:14; RTX161.

99. Rambus initiated licensing negotiations with Hitachi on October 22, 1999 by sending Hitachi a letter suggesting that Hitachi's products infringed Rambus's patents. HTX 015. Negotiations with Hitachi broke down when Hitachi refused to respond to Rambus's efforts to negotiate.

100. Rambus conducted a "beauty contest" in late November or early December 1999 to select counsel for litigation against Hitachi. Trial Tr. 1475:9-23. Rambus retained Gray Cary as litigation counsel for the Hitachi matter in late 1999. Trial Tr. 705:6-9.

101. Rambus filed suit against Hitachi on January 18, 2000.

102. Steinberg testified that in or around December 1999, he and Karp identified Rambus personnel who were likely to have relevant documents and told them to retain all such documents. Lawyers from Gray Cary, after the firm was retained in connection with the Hitachi litigation, also instructed dozens of Rambus employees that they needed to preserve all documents that could be relevant to the litigation. Trial Tr. 775:12-776:10.

M. Rambus's Resumption of the Destruction of Documents

103. The Hitachi litigation was settled on June 23, 2000. On that day, Rambus sent an infringement notice to Hynix (then Hyundai), another DRAM manufacturer, initiating Rambus's licensing strategy as to Hynix.

104. On or about that day, Attorney Vincent resumed purging his patent files, approving his secretary's selection from the Rambus prosecution files of documents to discard.

105. In December 2000, Rambus moved to a new office building. As part of the office move, Sure Shred disposed of materials that did not need to be moved from the old site to the new site. Rhoads of Sure Shred explained that shredding in connection with office moves is one of the primary reasons his company is called because it is cheaper to shred than to move materials and destroy them later. Trial Tr. 1623:5-15.

106. Prior to the current litigation with Hynix, Rambus had not disclosed the *1057 amount of document destruction it had undertaken. Sean Cunningham, one of the Gray Cary attorneys on the Hitachi litigation team, was not aware of the Shred Days. Cecilia Gonzales, one of the Howrey Simon attorneys on the Hitachi and Infineon litigation teams, was told by Attorney Steinberg about Shred Day 1998, but only found out about Shred Day 1999 and the 2000 destruction of documents after reading the Federal Circuit decision in the appeal from the original judgment in Rambus v. Infineon.

N. Effect of Document Destruction on Current Litigation with Hynix

107. Rambus kept no record of the documents it destroyed. Trial Tr. 334:2-12; 524:22-525:1; 1118:15-19. Karp expected everyone at Rambus to follow the document retention policy in discarding documents. Trial Tr. 334:13-17.

1. JEDEC-Related Documents at Rambus

108. Rambus's JEDEC representative Richard Crisp destroyed materials that he had received from JEDEC, such as unopened envelopes containing paper ballots and official minutes of JEDEC meetings. These documents, however, are publicly available from JEDEC.

109. After Crisp had left Rambus and after patent litigation had commenced, Crisp located various JEDEC trip reports and other JEDEC-related e-mails on a Macintosh drive in his attic. Upon locating these e-mails, Crisp turned them over to Attorney Steinberg for production in litigation. Steinberg then turned these emails over to outside counsel for use in the various lawsuits, including the instant case. Rambus appears to have produced to Hynix a complete set of Rambus's JEDEC trip reports. Trial Tr. 756:8-757:14; 1213:15-1215:10; HTX 182; RTX 206.

110. Although there could have been internal e-mail correspondence at Rambus about these trip reports which was destroyed pursuant to Rambus's document retention policy, such correspondence likely would have included Crisp as an author or recipient. Crisp produced the contents of his mailbox, after the Infineon litigation had commenced in January 2000. That mailbox contained Crisp's trip reports and other e-mails related to JEDEC. RTX 359. Crisp had specifically retained his JEDEC e-mails. Trial Tr. 121:15-1213:14. Therefore, with the trip reports and Crisp's e-mail, it appears that essentially the material JEDEC materials are available.

2. Contracts Between Hynix and Rambus

111. A "fairly complete record" has been produced of the external communications between Rambus and Hyundai regarding the negotiations for proposed amendment to their December 1995 RDRAM agreement. Trial Tr. 982:23-983:3.

112. The external correspondence between the parties regarding the Hyundai-LG merger and adoption of the LG Semiconductor licensing terms has been preserved and produced. Trial Tr. 983:10-24.

114. Rambus has either preserved and produced or logged as privileged numerous internal communications regarding these topics, including numerous internal communications discussing possible elimination or modification of the "Other DRAM" clause contained in the 1995 agreement and internal documents regarding the consequences of the Hyundai-LG merger. These communications include the following:

(a) an internal e-mail from Tate to Roberts dated April 2, 1998, discussing Hyundai and the possibility of "terminating their royalty deal for non-Rambus DRAMs" (Trial Tr. 1031:10-1032:23);
*1058 (b) an internal e-mail dated April 3, 1998, incorporating and replying to the April 2, 1998 Tate e-mail (Trial Tr. 1032:24-1034:1);
(c) an internal document dated April 3, 1998, discussing Hyundai and that Rambus's "action would most likely force them back to the negotiating table at which time we could argue to change some of the favorable terms we granted them first time around" (Trial Tr. 1047:2-1048:19);
(d) an internal e-mail from Mr. Mithani to executives at Rambus discussing Tate's April 2, 1998 directive, and noting that Rambus could "negotiate the special arrangement for allowing [Hyundai] to use our IP for the competing devices" (Trial Tr. 1048:20-1050:3);
(e) an internal e-mail reporting on a meeting with Mr. Yoo of Hyundai, and noting that Yoo was informed that one option was to "renegotiate the agreement and remove the Other DRAM clause" (Trial Tr. 1050:4-1051:2);
(f) an internal e-mail from Tate to Roberts dated July 10, 1998, discussing the amendment negotiations with Hyundai (HTX087; Trial Tr. 1029:1-20);
(g) an internal e-mail written by Tate regarding his discussion with K.H. Oh of Hyundai, regarding Rambus's "request for [Hyundai] to take goodie out of the contract," and reflecting Tate's agreement to "hold off on our concurrent/contract issues" pending agreement from Hyundai on marketing commitments (Trial Tr. 1051:3-1052:5);
(h) an internal meeting report by Mithani stating that Hyundai had a "change in attitude" and appeared to be "100 percent behind Rambus" in marketing (Trial Tr. 1055:3-1056:11);
(i) an internal e-mail from Steinberg to various Rambus executives dated January 4, 2000, "providing legal advice regarding assignment of license agreement" (HTX 216 (Rambus Privilege Log Entry 963)); and
(j) an internal memorandum from Mr. Bernsten to Steinberg dated January 27, 2000, "reflecting legal advice and analysis regarding Hyundai's licensing obligations to Rambus" (HTX 216 (Rambus Privilege Log Entry 4048)).

3. Patent-Related Documents

114. Rambus has produced various documents relating to the conception and reduction to practice of the patented inventions prior to the April 1990 filing date, of the '898 application. Trial Tr. 930:2-9. These include inventor notes (Trial Tr. 932:24-933:15; 935:12-16; 1003:2-23), computer simulations (Trial Tr. 1004:19-1005:19), drafts of patent applications (Trial Tr. 1045:22-1046:4), and technical presentations given by the inventors. Trial Tr. 1046:20-24.

115. Rambus has produced several boxes of prior art. Trial Tr. 949:17-19. The prior art produced by Rambus includes prior art references with fax lines showing that the art was ordered in 1996, prior art references with a few portions underlined, and prior art references containing a few other sorts of notations. Id. at 1009:17-1022:13. The evidence does not support a contention that Rambus reconstructed prior art that had been destroyed in advance of litigation. Further, no evidence was offered to show any prior art that any of Rambus's litigation opponents has identified as material and not cited by the inventor to the Patent Office in connection with the Farmwald/Horowitz line of applications. See Trial Tr. 1786:24-1787:20.

116. The evidence does not show that Rambus targeted or destroyed prior art pursuant to its document retention policy. All witnesses who testified on the matter confirmed that they had retained prior art *1059 and that they did not destroy it in implementing the Document Retention Policy. The evidence also does not show that infringement analyses or reverse engineering documents were targeted or destroyed as a result of Rambus's Document Retention Policy. The evidence established that Rambus conducted infringement analyses for the patents-in-suit after they issued in the 1999 and 2000 time frame, and that these documents were maintained under the Document Retention Policy and produced (or logged) in this case.

117. Jose Moniz began working on Rambus matters at Neil Steinberg's law firm in Virginia in January 1999. Trial Tr. 1626:15-1627:3; 1628:7-12. In late September or early October of 1999, Moniz became a Rambus employee. Trial Tr. 1626:10-12.

118. As part of his duties while working for Steinberg in early 1999, Moniz conducted a search for prior art relating to Rambus patent applications claiming priority to the original '898 application. Trial Tr. 1629:12-25.

119. Moniz obtained copies of any prior art that he believed might be material to one or more of Rambus's patent applications. This prior art was all submitted to the PTO. Trial Tr. 1631:23-1632:6; 1637:9-23.

120. Moniz added copies of the prior art that he found in his search to a collection of prior art that had already been gathered and stored the combined collection in banker's boxes. Trial Tr. 1632:7-18; 1638:16-25.

121. When Moniz joined Rambus, the boxes of prior art that he had maintained in Virginia were shipped to Rambus where Moniz continued to maintain them until two or three years ago when they were turned over to the litigation group. Trial Tr. 1639:15-25; 1658:5-9.

122. The prior art collected by Moniz in 1999 does not appear to have been destroyed, and copies of that prior art was produced to Hynix in this litigation. Trial Tr. 1641:13-1642:9; 1649:11-18; 1653:8-1658:4.

123. Moniz and Steinberg testified that they may have discarded duplicate copies of prior art containing notes related to their categorization of the art in order to determine to which of the Rambus patent applications the art could be material. Trial Tr. 1452:1-8; 1454:23-1455:4; 1641:3-12. Moniz also no longer has a spreadsheet, created and maintained while he was employed by Steinberg, logging the prior art collected for disclosure to the PTO. Trial Tr. 1659:13-25. However, the court concludes that no information was lost because the substance of the notes and spreadsheet was preserved in the Information Disclosure Statements submitted to the PTO, which show in which applications a particular prior art reference was cited. Trial Tr. 1481:7-1482:20; 1661:10-22.

II. CONCLUSIONS OF LAW

A. Focus of Dispute

The dispute between Rambus and Hynix focuses on the circumstances in which Rambus formulated, adopted, and implemented its document retention policy. Hynix argues that Rambus's contemplation of litigation against DRAM manufacturers triggered a duty requiring Rambus to retain all documents relevant to any potential patent litigation between it and any DRAM manufacturer. Hynix further asserts that since Rambus adopted its document retention policy and destroyed documents at the same time it was contemplating litigation, it spoliated evidence. This spoliation, claims Hynix, prejudiced its rights and warrants the application of the unclean hands doctrine and the dismissal of Rambus's patent claims.

*1060 B. The Defense of Unclean Hands

Under both federal and California law, one who has behaved inequitably or in bad faith is denied relief under the unclean hands doctrine. Precision Instr. Mfg. Co. v. Auto. Maint. Mach. Co., 324 U.S. 806, 814, 65 S. Ct. 993, 89 L. Ed. 1381 (1945), Jarrow Formulas, Inc. v. Nutrition Now, Inc., 304 F.3d 829, 841 (9th Cir. 2002); Fibreboard Paper Products Corp., 227 Cal. App. 2d 675, 726-29, 39 Cal. Rptr. 64 (1964). The unclean hands doctrine "closes the doors of a court of equity to one tainted with inequitableness or bad faith relative to the matter in which he seeks relief." Precision Instr. Mfg. Co., 324 U.S. at 814, 65 S. Ct. 993. The party seeking the application of the unclean hands doctrine generally bears the burden of proving each element. Omega Industries, Inc. v. Raffaele, 894 F. Supp. 1425, 1431 (D.Nev.1995) (citing Conan Properties, Inc. v. Conans Pizza, Inc., 752 F.2d 145, 150 (5th Cir.1985)). Spoliation of evidence can constitute the necessary "inequitableness or bad faith" for successful assertion of the unclean hands defense. See Samsung Electronics Co., Ltd. v. Rambus, Inc., 386 F. Supp. 2d 708 (2005). Further, if spoliation is shown, the burden of proof logically shifts to the guilty party to show that no prejudice resulted from the spoliation. The reason is that it is in a much better position to show what was destroyed and should not be able to benefit from its wrongdoing. See Anderson v. Cryovac, Inc., 862 F.2d 910, 925 (1st Cir. 1988); National Association of Radiation Survivors v. Turnage, 115 F.R.D. 543, 557 (N.D.Cal.1987).

C. Spoliation of Evidence

The court first addresses the question of whether Rambus spoliated evidence. "Spoliation of evidence is `the destruction or significant alteration of evidence, or the failure to properly preserve property for another's use as evidence in pending or reasonably foreseeable litigation.'" Zubulake v. UBS Warburg LLC, 220 F.R.D. 212, 216 (S.D.N.Y.2003) (quoting West v. Goodyear Tire & Rubber Co., 167 F.3d 776, 779 (2d Cir.1999)). "Defendants engage in spoliation of documents as a matter of law only if they had `some notice that the documents were potentially relevant' to the litigation before they were destroyed." United States v. Kitsap Physicians Service, 314 F.3d 995, 1001 (9th Cir.2002) (citing Akiona v. United States, 938 F.2d 158, 161 (9th Cir.1991)).

A legitimate consequence of a document retention policy is that relevant information may be kept out of the hands of adverse parties. See Arthur Andersen LLP v. United States, 544 U.S. 696, 125 S. Ct. 2129, 2135, 161 L. Ed. 2d 1008 (2005) ("`Document retention policies,' which are created in part to keep certain information from getting into the hands of others, including the Government, are common in business."). "It is, of course, not wrongful for a manager to instruct his employees to comply with a valid document retention policy under ordinary circumstances." Id. In contrast, however, a document retention policy adopted or utilized to justify the destruction of relevant evidence is not a valid document retention policy. See, e.g., Carlucci v. Piper Aircraft Corp., 102 F.R.D. 472, 486 (D.C.Fla.1984) (entering default against defendant who deliberately destroyed documents to prevent discovery, finding that the defendant "utterly failed to provide credible evidence that a [bona fide, consistent and reasonable document retention policy] existed"). It follows that implementing such a policy in advance of reasonably foreseeable litigation would not be proper and could constitute spoliation. The primary question before the court is whether Rambus adopted and implemented its document retention policy in advance of reasonably foreseeable litigation *1061 for the purpose of destroying relevant information.

D. Pending or Reasonably Foreseeable Litigation

Rambus initiated its first suit against a DRAM manufacturer, Hitachi, on January 18, 2000. Litigation was not actively pending when Rambus formulated and adopted its document retention policy. Thus, litigation was not pending at the time of the 1998 and 1999 Shred Days. The questions are, however, when litigation by Rambus was reasonably foreseeable and when a duty to preserve evidence arose.

"Sanctions may be imposed on a litigant who is on notice that documents and information in its possession are relevant to litigation, or potential litigation, or are reasonably calculated to lead to the discovery of admissible evidence, and destroys such documents and information." Wm. T. Thompson Co. v. General Nutrition Corp., 593 F. Supp. 1443, 1455 (C.D.Cal.1984) (emphasis added); see also Bayoil, S.A. v. Polembros Shipping Ltd., 196 F.R.D. 479, 483 (S.D.Tex.2000) ("Notice does not have to be of actual litigation, but can concern `potential' litigation. [citation] Otherwise, any person could shred documents to their heart's content before suit is brought without fear of sanction."). The Ninth Circuit has not expressly defined the term "potential litigation."

In the Second Circuit, the obligation to preserve evidence arises when "the party has notice that the evidence is relevant to litigation—most commonly when suit has already been filed, providing the party responsible for the destruction with express notice, but also on occasion in other circumstances, as for example when a party should have known that the evidence may be relevant to future litigation." Kronisch v. United States, 150 F.3d 112, 126 (2d Cir.1998). The American Bar Association's Civil Discovery Standards from August 1999 also provide some guidance. Standard No. 10 states, "When a lawyer who has been retained to handle a matter learns that litigation is probable or has been commenced, the lawyer should inform the client of its duty to preserve potentially relevant documents...." ABA Section of Litigation, Civil Discovery Standards, August 1999, Standard No. 10. The comments to this standard clarify that the "probable" language means that litigation must be more than a possibility, citing Iowa Ham Canning, Inc. v. Handtmann, Inc. 870 F. Supp. 238, 245 (N.D.Ill.1994) (noting that under Iowa law "[t]he requisite knowledge for imposing sanctions ... is not the `potential' for litigation, but the `contemplation' or `anticipation' of product litigation.") (citing State Farm Fire & Cas. Co. v. Frigidaire, 146 F.R.D. 160, 162 (N.D.Ill.1992)). Litigation "is an everpresent possibility in American life." Nat'l Union Fire Ins. Co. v. Murray Sheet Metal Co., Inc., 967 F.2d 980, 984 (4th Cir.1992).

Hynix correctly argues that in this case "probable" must be viewed from the perspective of a plaintiff, who is in control of when the litigation is to be commenced. This, according to Hynix, means that litigation is probable when litigation is contemplated. And, here, Hynix contends that Rambus contemplated litigation as of the time that it began to formulate a litigation strategy. In support of its argument, Hynix cites Silvestri v. General Motors Corp., 271 F.3d 583 (4th Cir.2001), in which the court found sanctionable spoliation where the plaintiff caused a car that was the subject of a product liability suit to be unavailable for litigation prior to filing suit. There are, however, significant factual distinctions between Silvestri and the present case. In Silvestri, there was a single piece of critical evidence, the damaged car. Plaintiff knew he was going to make a claim, had experts inspect the car for litigation *1062 purposes, had hired litigation counsel, and had even been advised by one of his experts that the car should be retained because the defendant would want to inspect it. The plaintiff nevertheless allowed the car to be destroyed.

Here, by contrast, the path to litigation was neither clear nor immediate. Although Rambus began to plan a litigation strategy as part of its licensing strategy as early as February 1998, the institution of litigation could not be said to be reasonably probable because several contingencies had to occur before Rambus would engage in litigation: (1) the direct RDRAM ramp had to be sufficiently developed so as not to jeopardize RDRAM production; (2) Rambus's patents covering non-RDRAM technology had to issue; (3) product samples from potentially infringing DRAM manufacturers had to be available in the market; (4) the non-compatible products had to be reverse engineered and claim charts made showing coverage of the actual products; (5) Rambus's board had to approve commencement of negotiations with a DRAM manufacturer; and (6) the targeted DRAM manufacturer had to reject Rambus's licensing terms.

Rambus planned to begin its licensing strategy only at the time the DRAM manufacturers were locked in to RDRAM production. By October 1998, the projected time frame for this was early 2000. Although Karp's October 1998 presentation projected that Rambus might be able to demonstrate that Mosel and Nanya had SDRAM products that directly infringed a pending access time register patent by the first quarter of 1999 and thus Rambus could potentially state contributory infringement and inducement claims against companies like Acer, SIS, and VIA for SDRAM and DDR, the presentation also recommended not even initiating licensing negotiations.

▄ DO NOT ROCK THE DIRECT BOAT.
▄ We should not assert patents against Direct partners until ramp reaches a point of no return (TBD)
▄ Probably not until Q1'00
* * *
▄ However, the Big Question Is— WHAT'S THE RUSH?
▄ What is the compelling business reason? I can't think of any.
▄ Keeping the Maytag repairman busy is not a valid reason
▄ IMHO, risks of damaging establishment of dominant standard outweigh potential return
▄ Lets not snatch defeat from the jaws of victory

HTX 128.003. The recommendation not to go forward with licensing against potential infringers at the time was consistent with Rambus's desire to ensure that DRAM manufacturers fulfill Intel's RDRAM requirements.

The patents Rambus asserts in this case did not begin to issue until June 22, 1999. Hynix nevertheless contends that Rambus had issued patents that it could have asserted against DRAM manufacturers as of January 13, 1998, the day it began meeting with Cooley attorneys. As of that date Rambus had identified to Cooley the '481 and '327 patents as "two key patents." HTX 376. In the Nuclear Winter Scenario memo, written at the end of 1998 or beginning of 1999, Rambus hypothesized how to convince Intel to continue their dealings in the event that the relationship between Intel and Rambus was compromised by the existence of another technology. This memorandum suggested that three Rambus patents could be leveraged to demonstrate that Rambus possessed intellectual property that covered DRAM solutions: '481, '327, and '580. Id. *1063 Karp testified that these patents were not strong patents and that Rambus preferred not to have to litigate them and that the hypothetical situation was set forth as a worst case scenario. Although there is no specific expression in the memorandum that the patents were potentially weak, the entire scenario was presented as a potential last resort tactic to convince Intel that RDRAM was still the best technology option and that Intel would not be able to completely walk away from Rambus's intellectual property by moving to other DRAM solutions.

Once customer samples were available in the market, Rambus planned to engage in reverse engineering to produce claim charts for use in its license negotiations. The Nuclear Winter Scenario memorandum demonstrates that, as of the beginning of 1999, customer samples of potentially infringing devices were not yet available and Rambus would, in the event of litigation, have to build claim charts based on datasheets rather than reverse engineering actual product. While the availability of customer samples does not appear to have been a necessary requirement for initiating licensing discussions, the fact that customer samples were not yet available supports the conclusion that, at least as of early 1999, Rambus was not contemplating initiating either licensing negotiations or litigation at that time.

Approval by Rambus's board to begin negotiation with DRAM manufacturers had not been obtained as of August 1999. Furthermore, and even more telling, Rambus had not budgeted for litigation. As of the June 1999 budget presentation, Rambus had budgeted for continued reverseengineering, and for continued prosecution of its patents, but had not budgeted, the $1.5 to $3 million projected to pursue litigation.

Rambus did view in early 1998 that future litigation against some DRAM manufacturer was possible if licensing negotiations failed. This expectation of involvement in litigation in this particularly litigious field was probably similar to that held by others in the industry who either anticipated the possibility of suing or being sued. Although Tate insisted that Rambus really wanted to avoid litigation when he and Karp began meeting with the Cooley attorneys, the evidence presented demonstrates that, once license negotiations commenced, Rambus wanted to "go in quickly and proceed either to a license or litigation." HTX 395. Rambus was also "looking for licensing rate that tells them it costs to infringe." Id. And Tate acknowledged that the 5% royalty Rambus was contemplating in early 1998 was "challenging" and that manufacturers might tell Rambus to "pound sand." Trial Tr. 1290:16-1291:1. In light of the direction given to the Cooley attorneys and the bullet item in Geoff Tate's July 1998 goals ("license OR sue"), HTX 094, Tate's testimony that the last thing he wanted was litigation with a major company reflects a less aggressive approach than the contemporaneous documents suggest. Nevertheless, it does appear that Rambus did not actually intend to initiate licensing negotiations for noncompatible users until certain contingencies occurred, which did not happen until late 1999.

Karp's IP Strategy Update dated September 24, 1999 reflects the turning point in Rambus's litigation intentions. In that document, Rambus acknowledges that "Intel has already started to let go." HTX 244.002. The presentation clearly states that Rambus's intellectual property in its patents must be substantiated by either settlement with "an industry powerhouse" or "winning in court," and acknowledges that some manufacturers will never settle. Id. At this point, Rambus appears to be *1064 ready to seriously consider actually filing suit against someone. Rambus's concerns about the direct RDRAM ramp were now outweighed by the prospect of losing Intel business; Rambus's patents had begun to issue; product samples from DRAM manufacturers were available in the market; Rambus was in the process of reverse engineering those products and producing claim charts; Rambus's board was now being presented with the necessity of launching the licensing and litigation strategy; and there was an explicit acknowledgment that at least some targeted DRAM manufacturers would probably reject Rambus's licensing terms. Notably, this presentation occurred after the Shred Day in August 1999.

In sum, although Rambus began formulating a licensing strategy that included a litigation strategy as of early 1998, Rambus did not actively contemplate litigation or believe litigation against any particular DRAM manufacturer to be necessary or wise before its negotiation with Hitachi failed, namely in late 1999. While hiring of litigation counsel or actually filing suit is certainly not necessary to demonstrate that a company anticipates litigation, in light of the record presented, it would appear that litigation became probable shortly before the initiation of the "beauty contest" in late 1999 in which litigation counsel for the Hitachi matter, Gray Cary, was selected. Thus, Rambus's adoption and implementation of its content neutral Document Retention Policy in mid-1998 was a permissible business decision. The destruction of documents on the 1998 and 1999 Shred Days pursuant to the policy did not constitute unlawful spoliation.

Rambus did shred additional records when it moved its offices in December 2000. Hynix initiated the instant lawsuit against Rambus on August 29, 2000. Therefore, Rambus was engaged in litigation at the time of its move and attendant disposal of documents. This raises the question of whether the third shredding event constituted spoliation. The evidence presented by Hynix did not demonstrate that any documents material to Rambus's patent claims were destroyed in conjunction with the 2000 move. At best, Hynix speculated that internal documents relating to the 1998 negotiations between Rambus and Hynix over the "Other DRAM" provision or the Hynix-LGS post-merger license had been destroyed to the extent those documents survived the shred days in 1998 and 1999. Rambus has negated that speculation by producing or logging as privileged numerous internal communications regarding these topics.

The fact that Rambus has previously claimed work product protection for some documents dated prior to late 1999 does not dictate a finding that Rambus was anticipating litigation at the time the documents were created. A reference to "work product" on a privilege log does not support the conclusion that litigation was anticipated because the log was prepared by California lawyers. California law differs from federal law in that it protects a lawyer's work product prepared "in a nonlitigation capacity." County of Los Angeles v. Superior Court, 82 Cal. App. 4th 819, 833, 98 Cal. Rptr. 2d 564 (2000). Thus, "[t]he protection afforded by the privilege [under California law] is not limited to writings created by a lawyer in anticipation of a lawsuit. It applies as well to writings prepared by an attorney while acting in a nonlitigation capacity." Id. Additionally, Gray Cary attorney Sean Cunningham testified that his understanding of the California work product doctrine at the time the Rambus privilege logs were prepared "was that it could apply even where anticipation of litigation wasn't necessarily present, that if it were the mental impressions of a lawyer, that it could be *1065 claimed as work product." Trial Tr. 765-66. Cunningham was not aware of any determination that his team of lawyers and paralegals who worked on the privilege logs made as to any "specific date and time" by which Rambus anticipated litigation against a DRAM manufacturer. Id. at 766.

The court concludes that Rambus did not engage in unlawful spoliation of evidence. This conclusion does not mean that a party can destroy documents with impunity prior to contemplation of actual litigation. The implementation of a document retention policy that was intentionally designed to discard damaging documents should litigation later become probable or actually commence would be improper. The evidence here does not support the conclusion that Rambus intentionally designed its Document Retention Policy to get rid of particular damaging documents.

E. Additional Requirements for Application of Unclean Hands Defense

1. Bad Faith Requirement

Although Rambus did not unlawfully spoliate evidence, it may have shredded relevant documents. The destruction of relevant documents, without more, does not support an unclean hands defense. "Bad intent is the essence of the defense of unclean hands." Dollar Systems v. Avcar Leasing Systems, 890 F.2d 165, 173 (9th Cir.1989). Accidental, inadvertent, or even grossly negligent behavior does not support the application of unclean hands. Id.; cf. Telectron, Inc. v. Overhead Door Corp., 116 F.R.D. 107, 131 (S.D.Fla. 1987) ("Resort to `the most severe in the spectrum of sanctions'—dismissal of an action or the entry of default judgment—is clearly appropriate where the offending party is found to have exhibited `flagrant bad faith' and its counsel has acted with `callous disregard' for the rules of discovery.") (citing National Hockey League v. Metropolitan Hockey Club, 427 U.S. 639, 643, 96 S. Ct. 2778, 49 L. Ed. 2d 747 (1976)). The unclean hands defense applies "only if the court is left with a firm conviction that the defendant acted with a fraudulent intent in making the challenged claims." Jarrow Formulas, 304 F.3d at 842.

Hynix argues that Rambus adopted its Document Retention Policy intending to destroy evidence relevant to potential patent litigation and breach of contract claims. It also asserts that Rambus was aware of potential antitrust counterclaims and the possibility of an equitable estoppel defense based upon its non-disclosure to JEDEC members of its patent activities. With regard to patent-related materials within Rambus, the document retention policy and Karp's presentation emphasized that engineering notebooks and other patent-related documents were important to keep. Karp's presentation encouraged keeping technical meeting minutes, technical presentations, problem summaries, and the like.

With regard to JEDEC materials, Karp testified that he never thought that Rambus had any JEDEC issues because he had attended JEDEC meetings on behalf of Samsung and believed that Rambus was an "open book" at JEDEC. Trial Tr. 504:5-505:3. Moreover, Karp never thought there was an obligation at JEDEC to disclose patent applications, and he never saw one disclosed. Trial Tr. 505:4-11. In fact, Karp testified that the first time he became aware that someone might assert an equitable estoppel defense based on JEDEC was when he saw Hitachi's counterclaims. 506:11-16. Although this testimony is difficult to believe given that Karp was specifically warned about equitable estoppel and had some prior litigation experience involving it, Gordon Kelley, Chairman of the JEDEC committee Rambus *1066 attended, stated in 1993 that JEDEC did not require the disclosure of patent applications and that his company, IBM, would not disclose them. Trial Tr. 1198:5-1199:5; RTX 312. Further, the Federal Circuit has held that Rambus was not obligated by virtue of its membership in JEDEC to disclose any of the patents or patent applications that were issued or filed while Rambus was a JEDEC member. See Rambus Inc. v. Infineon Technologies AG, 318 F.3d 1081, 1105 (Fed.Cir. 2003). Even though Karp may have had more concern about equitable estoppel than he was willing to acknowledge, the Document Retention Policy formulated by Karp did not single out JEDEC related documents for destruction. In fact, documents were produced that clearly show that Rambus was warned about the risks of equitable estoppel being applied because it participated in JEDEC and failed to disclose its potential patent coverage.

The document retention policy did include Rambus's explicit request in April 1999 to Attorney Vincent at BSTZ to conform outside counsel's patent files for issued patents to the PTO file wrapper. This practice resulted in the destruction of documents on the left side and center of the prosecuting attorneys' tri-fold application files. Hynix does not challenge the practice of conforming the files, merely the circumstances under which the request was made by Rambus. These discarded documents would have included attorney interviews and correspondence with Rambus employees as well as drafts of the application that were not submitted to the PTO. Attorney Johnson advised Rambus to initiate this practice, and Vincent was not supplied with guidelines by Rambus as to what sorts of documents to discard. However, his own practice at BSTZ included purging of files after patent issuance. There is no indication that Rambus had particular concerns regarding the contents of those files. Rambus just wanted Attorney Johnson's recommendation carried out. Prior art documents were retained.

2. Nexus of Discarded Claims with Patent Claims

"It is fundamental to [the] operation of the [unclean hands] doctrine that the alleged misconduct by the plaintiff relate directly to the transaction concerning which the complaint is made." Dollar Systems, 890 F.2d at 173 (quotation omitted) (first alteration in original). The Supreme Court has explained that courts

do not close their doors because of plaintiff's misconduct, whatever its character, that has no relation to anything involved in the suit, but only for such violations of conscience as in some measure affect the equitable relations between the parties in respect of something brought before the court for adjudication. They apply the maxim, not by way of punishment for extraneous transgressions, but upon considerations that make for the advancement of right and justice. They are not bound by formula or restrained by any limitation that tends to trammel the free and just exercise of discretion.

Keystone Driller Co. v. General Excavator Co., 290 U.S. 240, 245-46, 54 S. Ct. 146, 78 L. Ed. 293 (1933); see Jarrow Formulas, 304 F.3d at 841 ("[U]nclean hands does not constitute `misconduct in the abstract, unrelated to the claim to which it is asserted as a defense.'") (quoting Republic Molding v. B.W. Photo Utilities, 319 F.2d 347, 349 (9th Cir.1963)); United States v. United Energy Corp., 1987 WL 4787, at *13 (N.D.Cal. Feb. 25, 1987) (requiring "direct nexus" between alleged misconduct and claims at issue in litigation). Because the unclean hands defense requires that any alleged misconduct "must pertain to the very subject matter involved and affect the equitable relations between the litigants," the defense will not act to deny relief simply "because the plaintiff may have acted *1067 improperly in the past or because such prior misconduct may indirectly affect the problem before the court." Washington Capitols Basketball Club, Inc. v. Barry, 419 F.2d 472, 478-79 (9th Cir.1969) (quotation omitted).

To establish the required nexus, Hynix was required to prove more than the mere fact that Rambus destroyed documents, or even a large number of documents. The duty to preserve evidence, once it attaches, does not extend beyond evidence that is relevant and material to the claims at issue in the litigation. See Zubulake, 220 F.R.D. at 217-18 (recognizing that rule requiring retention of all documents upon recognition of threat of litigation "would cripple large corporations"). Hynix did show that some relevant documents were probably destroyed pursuant to the document retention policy. For example, Horowitz testified that some original slides of early presentations he had been saving for his children had been destroyed in the first Shred Day. Trial Tr. 611:13-25; 613:15-616:20. Horowitz explained, however, that these presentations had been made publicly and copies were available elsewhere. Id. It is also probable that Rambus's request that BSTZ conform the patent prosecution files resulted in the destruction, at minimum, of notes of inventor meetings and correspondence between Rambus and BSTZ attorneys regarding the patent applications. Thus, the court concludes that Rambus disposed of some documents pursuant to its Document Retention Policy that had a nexus to the patent claims at issue.

3. Prejudice to Hynix

An unclean hands defense further requires proof that the offending conduct materially prejudiced a party's ability to defend itself. See Republic Molding Corp., 319 F.2d at 349-50 (in applying the unclean hands defense, "the extent of actual harm caused by the conduct in question" is "a highly relevant consideration"); PenneCom B.V. v. Merrill Lynch & Co., Inc., 372 F.3d 488, 493 (2d Cir.2004) (unclean hands defense requires proof that the plaintiff "has injured the party attempting to invoke the doctrine"); Lawler v. Gilliam, 569 F.2d 1283, 1294 n. 7 (4th Cir. 1978) (denying unclean hands defense where defendant failed to prove that it had been injured by plaintiffs conduct because "[t]he party to a suit, complaining that his opponent is in court with `unclean hands' because of the latter's conduct in the transaction out of which the litigations arose, or with which it is connected, must show that he himself has been injured by such conduct, to justify the application of the principle to the case") (quoting J. Pomeroy, A Treatise on Equity Jurisprudence § 399); JTH Tax, Inc. v. H & R Block Eastern Tax Servs., Inc., 128 F. Supp. 2d 926, 949 (E.D.Va.2001) ("To establish the affirmative defense of unclean hands, a defendant must demonstrate ... plaintiffs conduct injured the defendant."), affirmed in part and rev'd in part on other grounds, 28 Fed.Appx. 207 (4th Cir.2002).

Although Hynix has made a showing that Rambus destroyed some relevant documents, Rambus established that adequate similar and material documents or classes of documents were not destroyed. The evidence showed that Rambus has produced to Hynix a large volume of relevant and material documents. According to the testimony of Hynix's counsel, Rambus has produced approximately 1.2 million pages of documents responsive to Hynix's discovery requests in this case. Moreover, for each category of documents material to the validity or enforceability of Rambus's patents that Hynix argued Rambus did not preserve and produce, Rambus has shown by clear and convincing evidence that documents in that category were in fact produced. Although Hynix *1068 complained that a notebook, models, and detailed diagrams authored by Dr. Farmwald that Farmwald himself described as being from early to mid—1988 in a September 1995—mail to Crisp were never produced, Rambus has produced various documents relating to the conception and reduction to practice of the patented inventions prior to the April 1990 filing date of the '898 application. These include inventor notes, computer simulations, drafts of patent applications, technical presentations given by the inventors, and several boxes of documents that had been in Farmwald's possession (and thus not subjected to Rambus's enforcement of its Document Retention Policy). Rambus also produced several boxes of prior art. The prior art produced by Rambus included prior art references with fax lines showing that the art was ordered in 1996, prior art references with a few underlines, and prior art references containing a few notations. No evidence suggested that material, non-cumulative prior art exists that Rambus has not produced. Thus, it does not appear that Hynix has been deprived of material, non-privileged, and non-cumulative documents.

Additionally, to the extent that documents such as notes of interviews with the inventors and draft responses to the patent examiner may have been discarded when Attorney Vincent conformed the issued patent files to the PTO file wrapper, those documents would probably be privileged and not been discoverable. See In re Spalding Sports Worldwide, Inc., 203 F.3d 800, 805-06 (Fed.Cir.2000); Advanced Cardiovascular Systems, Inc. v. C.R. Bard, Inc., 144 F.R.D. 372, 378 (N.D.Cal.1992) ("[W]e believe that inventors and their patent lawyers often engage in quite substantial private dialogue as part of the process of shaping and focusing a patent application, and that it is reasonable for them to expect that dialogue to remain confidential."); McCook Metals L.L.C. v. Alcoa Inc., 192 F.R.D. 242, 252 (N.D.Ill.2000) ("A draft necessarily reflects the communications between a client and his attorney as the attorney attempts to put forth the invention in the best light possible to protect a client's legal right.").

With regard to the prior license agreement between Rambus and Hyundai, Hynix acknowledges that a "fairly complete record" has been produced of (1) the external communications between the parties regarding negotiations between Rambus and Hyundai in or around July 1998 for a proposed amendment to the parties December 1995 RDRAM agreement and (2) external correspondence between the parties regarding the Hyundai-LG merger and adoption of the LG Semiconductor licensing terms. Rambus has either produced or logged as privileged numerous internal communications regarding these topics, including numerous internal communications discussing possible elimination or modification of the "Other DRAM" clause contained in the 1995 agreement and internal documents regarding the consequences of the Hyundai-LG merger.

The court concludes that Hynix has not been prejudiced by the destruction of Rambus documents.

F. No Basis for Dismissal

Default and dismissal are proper sanctions for willful destruction of documents and records that deprive the opposing party of the opportunity to present critical evidence on its key claims to the jury. See, e.g., Professional Seminar Consultants, Inc. v. Sino American Technology Exchange Council, Inc., 727 F.2d 1470, 1474 (9th Cir.1984). "Although a particularly severe sanction, outright dismissal of a lawsuit is within the court's discretion." Aptix Corp. v. Quickturn Design Sys., 269 F.3d 1369, 1378 (Fed.Cir. 2001) (citing Chambers v. NASCO, Inc., *1069 501 U.S. 32, 44, 111 S. Ct. 2123, 115 L. Ed. 2d 27 (1991)). However, the Supreme Court has cautioned that "inherent powers must be exercised with restraint and discretion." Chambers, 501 U.S. at 44, 111 S. Ct. 2123. Dismissal is a harsh penalty and should be imposed only in extreme circumstances. See Ferdik v. Bonzelet, 963 F.2d 1258, 1260 (9th Cir. 1992).

Here, the court does not find dismissal to be an appropriate sanction because it does not find the application of the unclean hands doctrine to be warranted. Further, the evidence presented does not bear out Hynix's allegations that Rambus adopted its Document Retention Policy in bad faith. The evidence also does not demonstrate that Rambus targeted any specific document or category of relevant documents with the intent to prevent production in a lawsuit such as the one initiated by Hynix. The evidence here does not show that Rambus destroyed specific, material documents prejudicial to Hynix's ability to defend against Rambus's patent claims. Therefore, Hynix's unclean hands defense fails.

NOTES

[1] Hynix was referred to as "Hyundai" prior to Hyundai's merger with LG Semiconductor in 1999.

[2] "HTX ___" refers to Hynix's trial exhibit number___; "RTX___" refers to Rambus's trial exhibit number___.