Why Hynix Will Lose The Evidentiary Hearing . . . Guest Analysis By James Kelley, Ph.D.

November 11, 2005

Copyright 2005, Kelley and Associates, LLC, All rights Reserved

(Posted on Treowth with permission of the author. Format revised for posting.)

****
PREFACE
This is an attempt to predict the verdict in the HYNIX versus RAMBUS Evidentiary Hearing and to give the reasons for that verdict. The author is not an attorney and is not an investment advisor and the use of this information should be measured against other sources and expert opinion.

It is possible that Judge Whyte may find some other legal standard to measure RAMBUS conduct against. (I personally believe that is not likely but there is nonetheless always the possibility.) In view of what happened in Judge Payne’s court in 2004 there is always the possibility of an unexpected outcome. So be very careful!

By accepting and reading this report you are acknowledging full responsibility for all your own investment decisions.

This document has drawn on the RAMBUS and HYNIX “findings of fact” and also the legal briefs which were filed with the court prior to the beginning of the Evidentiary Bench trial as well as the detailed notes of the Evidentiary Hearing.

If you wish more detail there is 156 page document that has the detail of the Judge Whyte Evidentiary Hearings.

This can be purchased for a modest sum at

rblegalreport@sbcglobal.net

This material is copyrighted to limit the distribution and while discussion of this material is encouraged topically copying material from the report for public dissemination is not allowed for good reasons.

Thank you and good luck!

(Page 1)

****

Table of Contents

PREFACE. 2
INTRODUCTION.. 3
The HYNIX CHARGES. 4
THE EQUITABLE ESTOPPEL DEFENSE. 5
SUMMARY OF EQUITABLE ESTOPPEL EVIDENCE: 8
AN EQUITABLE ESTOPPEL VERDICT: 8
THE SPOLIATION/UNCLEAN HANDS DEFENSE. 9
HYNIX MUST PROVE THE FOLLOWING: 9
THE CRITICAL QUESTIONS FOR THE COURT: 9
WHEN DID RAMBUS DUTY TO PRESERVE EVIDENCE BEGIN?. 9
DID RAMBUS WILLFULLY AND INTENTIONALLY DESTROY EVIDENCE?. 11
THE HYNIX NEXUS. 11
SUMMARY of UNCLEAN HANDS PROOF: 13
AN UNCLEAN HANDS VERDICT: 13

(Page 2)

****

INTRODUCTION
Hynix has painted a picture of Rambus as a company intent on litigating its non-compatible patent portfolio starting as far back as 1993. The scope of inquiry in this trial has largely been limited to the 1998, 1999 and early 2000 period by Judge Whyte.

This is the period in which Rambus first started up its patent licensing program for SDRAM. It is important to keep in mind that there were no commercial parts available for DDR or GDR during this period and SDRAM was just becoming available commercially. It is also important to keep in mind that RAMBUS only had a couple of non-compatible patents issued in 1998 and these patents were not considered strong enough on which to base a licensing program. It was not until the later half of 1999 that Rambus patent portfolio was strong enough to be used in a licensing program. Finally, it is important also to remember that RAMBUS at this time was focusing most of its engineering resources on the launch of Direct RDRAM, which had been chosen as the PC platform main memory by INTEL in the later part of 1996.

This charges against Rambus are that it destroyed not just documents but actually destroyed evidence related to prior art on the patents, that it destroyed emails essential to Hynix’s defense, that it negotiated in bad faith and that its employees lied under oath.

Hynix is claiming that Rambus document destruction has “prejudiced” their ability to defend themselves not only in this “spoliation” hearing but also in the infringement trial

This picture Hynix is drawing is at odds with commonsense. At the time, Rambus believed that their success was assured with the launch of Direct RDRAM on the Intel PC platform. There was no need to license SDRAM, etc if RDRAM was a success.

Moreover, a small (150 person) company would not attempt to manipulate, trick or defraud a 25 billion dollar DRAM industry. It defies common sense. Anyone of those companies has more lawyers than RAMBUS has employees. Yet this is exactly what Hynix is claiming. They are claiming that Karp was the mastermind of all this and that he kept his lawyers in the dark about his plans. Last year it was Dan Johnson and Karp masterminding this scheme.

(The HYNIX “conspiracy theory” was severely weakened by the HYNIX attorneys’ admission to Judge Whyte that they did not see anything wrong with Daniel Johnson’s advice to RAMBUS regarding a document retention program.)

(Page 3)

****

Throughout this hearing there has been an attempt on the part of HYNIX attorneys to “smear” RAMBUS witnesses like Mr. Tate, Mr. Crisp, Mr. Karp, Mr. Steinberg, Mr. Vincent, and even Mr. Johnson. The purpose of the “smear” is to get RAMBUS testimony discounted in favor of HYNIX. While this strategy worked in Judge Payne’s court, it is known that Judge Whyte does not like this sort of defense. It seems to work against HYNIX throughout this hearing.

THE HYNIX CHARGES
Hynix is making two broad defenses to the patent infringement charges against it as follow:
A) Rambus knew that they might be barred by equitable estoppel and

B) Rambus destroyed documents before its planned litigation against the DRAM manufacturers to obstruct their defenses.

These broad defenses will be addressed one at a time.

Of the two defenses (B) the so-called SPOLIATION/UNCLEAN HANDS defense must be taken very seriously because Judge Payne has ruled from the bench that RAMBUS “spoliated” and barred them from enforcing the patents in suit in his court after a three-day bench trial in 2004. The 2004 spoliation case was based on a “conspiracy theory” which had Dan Johnson and Joel Karp masterminding the destruction of evidence as they planned to sue the DRAM manufacturers.

This year HYNIX has dropped Dan Johnson as one of the masterminds and has left Joel Karp as the sole mastermind.

THE EQUITABLE ESTOPPEL DEFENSE
To prevail with this equitable estoppel defense HYNIX must prove that RAMBUS intentionally misled JEDEC and others into thinking that RAMBUS would not enforce their patents.

RAMBUS had no duty to inform anyone of their patent applications as a practical business matter and of course, actual patents are documents that inform the public.

HYNIX cannot prove this because all the memory manufacturers were already on notice of RAMBUS patents and the likelihood that those patents would eventually obtain broad coverage and the CAFC has already ruled that RAMBUS had no patents that read on the JEDEC standard while at JEDEC. Also, RAMBUS responded to competitive threats by directly warning them of possible patent infringement.

Specifically the first HYNIX “broad finding” is as follows:

A) RAMBUS was on notice that JEDEC members were likely to assert that its failure to disclose it IP claims barred enforcement of its patents against JEDEC standard SDRAM.

This is the shorter and easier of the two broad findings of fact. This concerns the equitable estoppel issue.

The sub heading here is as follow:

(Page 4)

****

1) Between 1992 and 1997, Rambus employees and executives were repeatedly informed by Rambus’s in-house and outside counsel that Rambus’s participation in JEDEC could be the focal point of an equitable estoppel defense in future patent litigations.

To support this line of argument Hynix cites the following exhibits:

HTX-216 Rambus privilege log 9/23/1995

HTX-002 Tate email re: JEDEC/Gordon

HTX-066 Crisp email to Diepenbrock and EXEC re: SyncLink (9/23/1995)

HTX-073 Letter from Vincent to Diepenbrock re: the FTC DELL consent Decree 12/19/1995
Vincent’s July12, 2001 Deposition by Micron

HTX-021 Handwritten notes by Vincent 1992

HTX-077 Handwritten notes by Vincent 1/11/1996

DEP Vincent’s April 12, 2001 Deposition by Infineon

DEP Vincent Infineon Deposition (3/14/2001)

HTX-078 Crisp Email to Tate re: avoidance of equitable estoppel (1/22/96)

DEP Vincent 4/11/2001 Infineon deposition

HTX-225 Crisp email to Mooring re: informing infringers

For this analysis, I will first select the relevant equitable estoppel selections from the evidentiary hearing in Judge Whyte’s court so that the reader can form his own opinion about this Hynix “finding of fact”. Next I will summarize what conclusions I think can be drawn from the evidence. Finally, I will give my “verdict” of whether this Fact was proved in court.

Evidence brought forth during Judge Whyte’s hearing:

Vincent’s testimony regarding Equitable Estoppel:

Vincent clarified the equitable estoppel issues, which he instructed Rambus.

Basically, if you say you will not enforce your patents and later you come back and attempt to enforce them you will face the strongest case of Equitable Estoppel. It is less clear when you remain silent and it is best not to vote at JEDEC.

(Rambus followed this advice they remained silent and did not vote at JEDEC.)
(Mr. Tate testified that in his visits to manufacturers he warned them that they might step on RAMBUS patents.)

(Page 5)

****

Vincent: “having trouble remembering what I said.” There could have been an equitable estoppel if Rambus had led JEDEC members into believing that they would not enforce their patent rights.

There is a discussion of downside risk versus upside potential. The risk as that someone might raise the equitable estoppel issue in the future. (I assume the upside could be the adoption of the RDRAM as a standard by JEDEC).

Lynch: At some point you developed a concern about equitable estoppel issues being raised concerning JEDEC membership and you advised Rambus not to participate. Did you participate in a telephone interview with Mr. Sikorsky, Mr. Pendarvis and Ms. ?
Vincent: I don’t recall. He can remember only Mr. Sikorsky.
At some point the DELL decision was discussed. Mr. Vincent does not recall whether the notes were made at the meeting or later.

(The DELL FTC issue resulted in a DELL patent becoming unenforceable because DELL remained silent during the standardization process. This case caused RAMBUS some concern but RAMBUS had no patents that READ ON JEDEC standards at the time.)

Vincent is not sure whether he advised them at that meeting or merely made some notes about the issue.

Vincent considered Mr. Crisp to be an engineer at Rambus.

Mr. Vincent does not recall the specific subject matter of the claims questions asked him by the engineers.

Mr. Karp asked him in April 1999 to purge his documents pursuant to the DRP.

Mr. Vincent testified that did not remember discussing any claims specifically directed at SDRAM. He remembered discussing the subject matter of the patents with the engineers. These discussions centered on specific features. The patent specifications were very long. He kept notes of those meetings.

Richard Crisp on Equitable Estoppel:
HTX-085 Whether to stop going to JEDEC Meetings (12/1992)

Email between Mooring and Crisp. Regarding equitable estoppel issue with respect to JEDEC.

Mr. Weed of Branson and Branson states that perhaps JEDEC requires patent disclosures but this is inadvisable.

(Page 6)

****

Mr. Crisp believes that he had a conversation with Lester Vincent in which Vincent told him patent disclosures were not required. (The CAFC ruled that this was the case.)

The testimony of Richard Crisp painted a picture of the time period between 1992 and the time he left the company. It reflects his concerns with attendance at JEDEC meeting and possible equitable estoppel issues. The company apparently got several different kinds of opinions on whether to continue at JEDEC or drop out it. Rambus appears to have followed the advice of Lester Vincent at Blakely Sokoloff in that it refused to comment on its patents and refused to cast ballots from 1992 onwards. Rambus had been unsuccessful in getting JEDEC to even let them propose RDRAM as a standard.
Rambus dropped out of JEDEC in late 1995 about the time the SyncLink competition was developing to attack RDRAM.

By 1994 Rambus was becoming acutely aware of the “copy cat” problems and was becoming interested in extending their patents to cover what they thought to be proposed products that would infringe properly drawn patents on their intellectual property. Unfortunately, Rambus was naïve about the difficulty of extending its patent coverage. They actually did not achieve coverage until 1998 almost 4-5 years later. Even then the 1998 patents turned out to be to poorly drawn so that they did not read strongly on SDRAM.

SUMMARY OF EQUITABLE ESTOPPEL EVIDENCE:
It seems clear from the evidence that Rambus attorney Vincent had some concern about Rambus attending JEDEC meeting in 1992-1993. Rambus mistakenly thought its RDRAM patent might cover works being done at JEDEC on SDRAM. This last fact was noted in the CAFC ruling. Specifically, they were not sure whether they had to disclose patents and patent applications to JEDEC.

Gordon Kelley who was the chairman at that time of JEDEC cleared up this matter inadvertently when he said that IBM would not disclose patents or patent applications that might read on JEDEC devices as a matter of policy.

Rambus appears to have followed Vincent’s advice without further concern. Then in late 1995 the FTC DELL decision came down and it placed a burden on a patent holder to reveal his patent if it read on a standard. Rambus had no patent that read on the standard at that time. A decision was made to not attend any more JEDEC meetings. Rambus paid no renewal fees on its JEDEC membership after that.

(Page 7)

****

The equitable estoppel issue came up in connection with a startup firm MOSYS which was formed by two former RAMBUS engineers and which was selling a product that looked like RDRAM. RAMBUS talked with this company and eventually licensed them.

The equitable estoppel issue surfaced once more in connection with SyncLink and the SLDRAM Consortium that was attempting to develop a competitor to RDRAM.

RAMBUS felt that it had to inform them of a probable infringement problem.

RAMBUS did inform them and there was a scandal relating to the Townsend and Townsend Law Firm that was representing Hyundai, the SLDRAM Consortium and RAMBUS all at the same time. Townsend and Townsend dropped the Consortium. The Consortium ultimately abandoned SLDRAM after being informed by Rambus that it would probably infringe. RAMBUS fired Townsend and Townsend. Townsend and Townsend still represents Hynix the successor company to Hyundai.

There was no further concern about equitable estoppel issues until the Hitachi and Hynix lawsuits in the year 2000.

AN EQUITABLE ESTOPPEL VERDICT:
Equitable estoppel is out of this case as a real issue.

The JEDEC equitable estoppel issues were addressed by the CAFC when it determined that RAMBUS had no duty to reveal either patents or patent applications to JEDEC. All other equitable estoppel issues were directly addressed by RAMBUS, i.e., MOSYS, SYNCLINK, and DELL. The record shows that RAMBUS had no further concerns about equitable estoppel issues for 4 years until the lawsuits which began in 2000.

One of the consequences of the loss of the equitable estoppel issue is that Rambus had no legal duty to preserve documents in connection with any of these equitable estoppel issues raised by Hynix. However, RAMBUS did preserve the documents anyway.

THE SPOLIATION/UNCLEAN HANDS DEFENSE
This issue divides itself into what Hynix has to prove and what the Judge has to decide. None of the witnesses gave any credible evidence of actual intentional and willful destruction of material evidence that could have been of any use to HYNIX in defending itself from patent infringement charges.

HYNIX MUST PROVE THE FOLLOWING:
1) RAMBUS failed to preserve evidence when it had a duty to retain evidence.

2) RAMBUS willfully and intentionally destroyed evidence.

3) The evidence destroyed was relevant and material.

4) RAMBUS conduct has prejudiced RAMBUS ability to defend itself in the current patent infringement suit.

(Page 8)

****

THE CRITICAL QUESTIONS FOR THE COURT:
1) When did RAMBUS duty to preserve documents begin?

2) Did RAMBUS willfully and intentionally destroy documents?

3) Was the evidence allegedly destroyed material and relevant to the patent infringement suit?

4) Has RAMBUS’ conduct prejudiced HYNIX’s ability to defend itself?

WHEN DID RAMBUS DUTY TO PRESERVE EVIDENCE BEGIN?
Hynix has taken the position that RAMBUS duty to preserve documents began at least 18 months before RAMBUS entered into litigation with Hitachi on the non-RDRAM patents. HYNIX in its case law references has claimed that the duty to preserve evidence begins at one of the following times:

When litigation begins;

When litigation could (reasonably) potentially happen; or

When litigation is anticipated.

HYNIX is arguing that RAMBUS was planning an aggressive litigation program at the beginning of 1998. Thus they are stating that RAMBUS had a duty to preserve relevant evidence at least to the beginning of 1998. Elsewhere they are arguing that this duty goes back to 1988.

(Page 9)

****

RAMBUS has adopted the more clear standard put forth by the American Bar Association in the year 1999. This standard presumably governed the actions of lawyers across the nation during that time period.

When litigation has commenced .

When litigation is not only anticipated but probable.

It is clear that RAMBUS followed the ABA standard.

By this standard, RAMBUS would have set the duty to preserve data in the time frame of November or December of 1999. Arguably, the date could have been set to October of 1999 when the board of directors approved the licensing program for Hitachi but Stone would argue that Hitachi could have settled and therefore a lawsuit was not probable at that time.

In any event there is no specific evidence that anything material to the patent infringement case that was destroyed during those three months. It is also clear that all the attorneys were acting as though the commencement of litigation was the time to put a “litigation hold” on the document retention program.

Rambus has presented a persuasive argument that the SDRAM, DDR patent licensing program was a backup plan to the RDRAM program with INTEL. If RDRAM had been successful that the SDRAM and DDR patent licensing program would have been unnecessary. During the timeframe of 1998 there was no solid reason to believe that RDRAM would not be successful.

HYNIX has tried to prove that RAMBUS was intending to initiate litigatation by charging very high royalties on SDRAM and DDR. HYNIX claims that RAMBUS was intending to charge 5% on these competing technologies.

In fact, this never happened during the HITACHI litigation. HYNIX failed to prove this assertion. RAMBUS never discussed its royalty rate with Hitachi at the first and final meeting before litigation. (In fact, SDRAM licenses are about .75% and DDR is well under 5% for
non-litigious licensees.)

DID RAMBUS WILLFULLY AND INTENTIONALLY DESTROY EVIDENCE?
HYNIX is asserting that RAMBUS willfully and intentionally destroyed not only documents but also evidence as part of a conspiracy. They have even implied that the outside lawyers were in on this conspiracy. In particular, Dan Johnson who provided the legal advice for establishing RAMBUS document retention/destruction program in 1998 is asserted to be part of the conspiracy.
HYNIX has taken the difficult position of asserting that document retention programs are perfectly legal but that somehow RAMBUS document retention program was just a smokescreen for destruction of evidence.

In order to prove this HYNIX must show that RAMBUS targeted specific material for document destruction and those targets contained relevant evidence. HYNIX cannot do this because RAMBUS did not target specific categories of documents in its document retention plan and it can be shown that RAMBUS made serious attempts to preserve evidence even when they could have gone undetected in destroying the evidence.

(Page 10

****
Because they cannot point to specific categories of evidence that were destroyed wholesale HYNIX has made much of the destruction of emails at RAMBUS.

The problem with this line of argument is that email was just becoming “discoverable” as evidence in 1998. Johnson advised RAMBUS to include an email policy as part of its document retention program. There was no duty to preserve email until later.

In order, for HYNIX to prove their case they must prove that RAMBUS intentionally destroyed evidence related to the patents in suit. It is not enough for them to show general document destruction. It is not enough for them to attempt to discredit the RAMBUS witness testimony from former hearing and trials. This is called the NEXUS and Judge Whyte has asserted that they must show the NEXUS between what was destroyed and the current patent infringement Lawsuit.

THE HYNIX NEXUS
HYNIX must show that the documents destroyed by RAMBUS are material and relevant to the patent infringement suit.

(I am omitting the JEDEC related document issues as irrelevant to patent infringement issues in this case.)

In order to make their case HYNIX attempted to use Theodore Brown a Townsend and Townsend lawyer and also a partner in the firm as an “expert witness”. Judge Whyte did not allow this as he said it would be “unseemly” to allow him to testify in that way. Mr. Brown was allowed only as a “fact witness”. HYNIX protested because Mr. Brown was the only “expert witness” in support of their Unclean Hands defense.

Judge Whyte ruled that Mr. Brown was a “biased” witness after Mr. Stone established his animus towards RAMBUS that grew out of Townsend and Townsend scandal in 1996 and 1997.

Mr. Stone caught Mr. Brown in one lie after another during cross-examination.

Mr. Brown listed the following as missing from RAMBUS patent related document production:
1) No patent related documents prior to 1990 were found in the document production.

2) No agreement between Farmwald and MIPS were found.

3) No agreement between the University of Illinois and Farmwald was found.

(Page 11)

****

4) A missing Farmwald notebook prior to 1990.

5) A 240-page gap in document production regarding the patents.

6) Mr. Moniz spreadsheet cross-referencing prior art with Rambus patents is missing.

7) Rambus used vendor supplied prior art in their production to HYNIX. HYNIX refers to this as “Salting the Prior Art production”.

8) “307 patent was not referenced by Rambus in its other patents.
(Brown was implying that RAMBUS committed fraud on the patent office by not disclosing the ‘307)

Mr. Stone eliminated 1 by showing Mr. Brown a “Spice” circuit simulation from Stanford dated prior to 1990 in the document production. Mr. Brown had testified there were none. Mr. Brown retorted that that was just “one example”. Mr. Stone called his bluff by stating that he had “more in the box”.

The issue of Farmwald’s employment at MIPS is not relevant to the validity of RAMBUS patents. Concepts are not patentable and it took Dr. Farmwald another two years to reduce his concept to the form of an invention. If in fact, this were a real issue, MIPS would rush in to assert its rights. Thus 2 is eliminated.

Mr. Stone produced the agreement between University of Illinois and Farmwald in which the University agreed not to assert patent rights over Farmwald’s inventions eliminating 3. Mr. Stone then asked Mr. Brown if he had done discovery of the University of Illinois. He said yes.

Mr. Stone also showed a page out of Farmwald’s inventor’s notebook that disposed of 4. When Mr. Brown retorted that it was just “one page” Mr. Stone said, “I have more in the box”. Mr. Brown acquiesced.

Mr. Stone then showed how HYNIX had come up with the 240-missing page gap in the document production. The documents were there. Again, Mr. Brown’s testimony was false.

Mr. Moniz spreadsheet was left on his old employer’s computer when he moved to RAMBUS in California. This is hardly a willful act and the cross-references are contained in the patents themselves so what is the problem? This disposes of 6.

The use of vendor supplied prior art is not a problem since as Mr. Stone pointed out “prior art always come from somewhere else” other than Rambus. It looks like RAMBUS ordered another patent search for production. But Rambus also produced the original prior art files HTX-524. “Prior art” is in the public record. RAMBUS conformed its patent files like other companies. This disposes of 7.

Mr. Stone shows the ‘307 patent being referenced in one of RAMBUS other patents showing that it was not intentionally left out of the prior art produced to the patent office. This disposes of 8.

(Page 12)

****

Every statement of “fact” made by Mr. Brown was discredited with the exception of the Excel spreadsheet that was left on a computer in Virginia. This is a secondary derived document and it cannot be argued that it has any importance in HYNIX defense or that its loss was in anyway “willful” or “intentional”.

SUMMARY of UNCLEAN HANDS PROOF:
HYNIX has failed to prove

1) That RAMBUS willfully and intentionally destroyed evidence.

2) That any of the documents destroyed were material and relevant to the patent infringement suit.

3) That RAMBUS’ conduct in anyway prejudiced HYNIX’s ability to defend itself.

HYNIX has also failed to prove that RAMBUS intended to sue the memory manufacturers beginning in 1998. In fact, litigation was just one of the many possible business decisions.

Thus, the date for “litigation hold” should be placed by the ABA guidelines no earlier than October 1999 if not December of 1999. HYNIX has failed to prove that any relevant or material documents were destroyed between October 1999 and January 2000.

There was no showing that prior to 1999 RAMBUS was under any duty to preserve emails that at that time were just becoming discoverable. It is also clear that the outside attorneys at that time were taking the decision to litigate or actual litigation as the time to place a “litigation hold” of document retention programs.
RAMBUS has shown HYNIX claims of spoliation to be speculative.

AN UNCLEAN HANDS VERDICT:
HYNIX loses the SPOLIATION/UNCLEAN HANDS Case due to a lack of any credible showing of willful or intentional destruction of any relevant and material documents related to the HYNIX defense against patent infringement. HYNIX chief witness was established to be biased and to have testified incorrectly on every major issue save one.

Judge Whyte- in my humble opinion- will use the ABA standard for defining the date of “litigation hold” because it was the standard in 1999.

(Page 13)