reprinted from http://finance.messages.yahoo.com/bbs?.mm=FN&action=m&board=4687909&tid=rmbs&sid=4687909&mid=710084 http://finance.messages.yahoo.com/bbs?.mm=FN&action=m&board=4687909&tid=rmbs&sid=4687909&mid=710085 http://finance.messages.yahoo.com/bbs?.mm=FN&action=m&board=4687909&tid=rmbs&sid=4687909&mid=710086 http://finance.messages.yahoo.com/bbs?.mm=FN&action=m&board=4687909&tid=rmbs&sid=4687909&mid=710106 http://finance.messages.yahoo.com/bbs?.mm=FN&action=m&board=4687909&tid=rmbs&sid=4687909&mid=710107 http://finance.messages.yahoo.com/bbs?.mm=FN&action=m&board=4687909&tid=rmbs&sid=4687909&mid=710108 Many thanks to waskilroyhere! *************************************************************************************************** FTC part 1 by: waskilroyhere 12/09/04 11:49 pm Msg: 710084 of 710116 thought I posted this, but don't see it: The commissioners, well worth a look if you haven't already: http://ftc.gov/bios/commissioners.htm The room is like a small courtroom, wood paneled, with an elevated bench across the front at which the five commissioners sit. There are seats for about eighty, including counsel, three or four for both parties, forty seats reserved for interested parties, Rambus in-house counsel, Mark Horowitz, Bloomberg reporter, and on the other side of the aisle, M. Sean Royall, and a number of FTC employees. About thirty seats were behind those. Among those in back were Stuart Steele, a number of other Rambus investors (many of whom no doubt post to the boards) and two Unocal attorneys, apparently there to get a sense of what the newly constituted commission is like. Geoffrey Oliver (CC) was complaint counsel. Greg Stone (GS) presented what he called the "fact intensive" and Douglas Melamed (DM) the "legal and policy intensive" parts for Rambus. Complaint counsel opened with the first hour, Rambus followed with a half hour GS, 15 min DM, 15 min GS, and then CC rebuttal of about 15 min. Hearing started at 10, ended about 12:15. My understanding is that there was a closed, executive session this afternoon, but I am not certain. What follows is my paraphrase that I think generally gets the gist of what was said. I think it would be really helpful to all of us if Nic, 12S, JA or another attorney would read over particularly the Melamed section where there are a lot of shorthand exchanges among lawyers who understand what the referenced cases and sections of law mean because I do not and the reportage suffers on account. CC:Unlawful method of competition is a violation of section 5. Impoetance of the case is to both consumers and SSO's. Should antitrust laws encourage or condemn practices that cause price increase not from market power but from secretly using SSO's for patenting purposes. Result will determine whether consumers pay up to $3B more and will determine behavior in SSO's. Will Antitrust laws allow companies to plot and twist SSO regulations or support SSO procompetitive regulations. Our focus is on exclusionary conduct: there is no dispute with regard to Bambus' behavior. No dispute about facts. Some facts are absent from ALJ's opinion. In December 1991, Rambus had no market power. Today it claims to extract royalties from DRAM and controllers. Their power is not derived from their failed DRAM but from subversion of a SSO. They joined Jedec in December 1991 and annually renewed. DM: Is the clarity of Jedec standards important? CC: Allied Tube is the important reference. DM: How do we give guidance? CC: Look to the purpose of the organization, its procedures and policies. DM: What if an organization had no policy? Could an argument still be made for anticompetitive harm? CC: The difference? A 1983 report by Muris stated that it is not appropriate for the agency to engage in rulemaking. Organizations knowledge of their industries is central to standards setting. The sign-in sheet had a statement about patent disclosure. It is clear they were concerned about disclosure of patents and applications. JL: Rambus disclosure of patents is an issue but what atout other members? CC: IBM notified of applications. (!!!) JL: Of sixty here only three applications. Rest are patents. So you think its clear everyone understood applications must be disclosed and did so? CC: Yes. JL: The ratio of patents to patent applications is 19:1. CC: There is no evidence of any company filing and later asserting against the standard. In three instances a company tried to assert and it caused an uproar. Richard Crisp saw the uproar over TI. JL: Five judges including three CAFC found no requirement. . . . CC: Standard was clear and convincing evidence. We showed Rambus did violate duty. Rambus was wrong in its assumptions about a need to show harm only to DDR standard. This is untrue. SDRAM affects subsequent standards. They were wrong about policy. Rambus assumes you must look for a start date. This is wrong. John Kelly was very clear. There is no date as to the start of discussion. DDR began in 1993. Crisp document from May 1993: G Kelly had announced SDRAM almost final, to be finished next week. On the left hand side of the page, Crisp's notes discuss votes on the standard and on the right the start of work on the next standard. One month later, Fred Ware, Rambus engineer, emails to Crisp and others working on amendments of applications to extend to other types of DRAM on chip DLL added to DDR standard. TL: If Rambus didn't mislead Jedec, then standard was adopted because of efficiency. CC: If Rambus disclosed, Jedec would have decided otherwise. TL: People who are misled have to be reasonable people. Shouldn't we apply this to Jedec members, too? CC: I respectfully disagree. They are, but they didn't have to be. We are here to represent the consumer. TL: We are here to determine whether Rambus misled Jedec. Did members act reasonably when they assumed Rambus wouldn't assert IP? JL: Rambus doesn't make anything. How could any reasonable person assume Rambus was ther for anything but to license their IP? CC: Jedec assumed they wouldn't assert beyond a narrow bus. (At this point he first employed an analogy between MM's with car being approached by Rambus with motorcycle. It struck me as a singularly distorted analogy. I don't think the commissioners responded to it and I neglected it. Others may have picked up on it and seen more significance.) DM: Suppose a weasely member waffles when asked? Do members have antitrust rresponsibility to find out. CC: Burden being on the patent holder is the simplest and least expensive solution to to getting information out. PH: May I direct your attention to lock-in? CC: Sure. PH: ALJ said lock-in was not a serious issue because of shifting market share of memory types. CC: They are all related. Companies know this and plan for it. AMD-Intel shows how difficult it is to go back and rip it out. PH: AMD performed five iterations in 1999-2003. Doesn't this refute lock-in? CC: Each of these was simply a refinement, not a total change. The need for tearing out a standard is what is disruptive. TL: Is there any evidence in the record about the amount of investment lost if reconfiguration is required? CC: Bechtelstein of Cisco said $1B for redesign. Brian Shirley of Micron said humdreds of millions. Others said it would be similarly expensive. DM: Return to reasonableness. Surely you're not saying they're not held to any standard. They must . . . to procompetitive standards.(Ref is to MM's here.) What if Rambus says, "I might, but I won't say"? Their duty is then to confront Rambus. This is important because you have said we can set the future behavior of SSO's. CC:What impact would if have had? If it would make a difference, the burden is on the patent holder, then the SSO. If the SSO gets it wrong, the patent holder should not be exonerated. DM: It's about drawing a line where the law steps in. CC: It is difficult, but the intention of the patent holder is being left out. Rambus conspired (my word to fill in here) at the highest level of the organization. OS: About expections, the participants in Jedec are reasonable people. They are sophisticated. They operate in an extremely competitive business. What did participants see as Jedec's purpose and what would they expect of members? CC: Inside and out of Jedec, people expected open standards available to all. PH: I notice your yellow light is on. Before your time is up, I'd like to change focus and talk about alternatives. If they were available, why did Rambus have to guarantee RAND and possibly see a shift away from their technology. CC: Between Crisp and Tate there was an exchange that said, "Why alert them to a patent issue they can reasonably work around?" PH: With RAND, would Jedec have had incentive to work around Rambus? CC: Yes, depending on royalty rate. It's clear some members would have liked to go with alternatives from the outset. PH: Do you believe CC must prove that for the section 2 violation? CC: The standard is Hovencamp (phonetic spelling, haven't a clue) and Microsoft, but even if Rambus proposed standards, Jedec would not have adopted absent RAND. There is evidence to support that. Specifically, NEC in 1997 had an issue with the '703 and Jedec the next month adopted an alternative. End of first hour Gregory Stone, in his words, "on behalf of Rambus, Drs. Farmwald and Horowitz." Stone to present fact intensive portion, Melamed to present legal and policy intensive portions. Seven issues and respond to questions. Over fifteen years ago, Farmwald and Horowitz confronted memory bottleneck with revolutionary solution on April 18, 1990. Forty-one patents had arisen from it by 1993. There is no complaint about how patents were obtained but complaint counsel wants to take away those patent rights because of claimed misrepresentations. DM: Why does antitrust law care? GS: No interest by itself. DM: But in the context of a SSO? GS: If there's a tort, a fraud against the organization. DM: If done properly, SSO is procompetitive and that is the interest of antitrust, so if a member has thwarted procompetitive purpose, and many would aruge that a hold-up does that, then antitrust should care. GS: Different organizations have different rules. The is a need to balance competing interests. Jedec never searched to see if patents were out there that applied. Jedec was happy just to hear from members if their proposals were based on their own patents. PH: Jedec rules were to prevent a takeover of a technology> GS: 1. McAfee, complaint counsel's witness, said if Rambus never joined there would be no change in Jedec; 2. Rambus was such revolutionary technology it would still have been adopted . . . rand. Willi Meyer's testimony acknowledged Rambus covers SDRAM. Gordon Kelly's and Mark Kellogg's notes corroborate this. TL: If it would have go Rambus' way anyway, why not make it all open and avoid the litigation? GS: Good reasons. Rambus attorneys of the time advised them to keep their patent applications private because they didn't know what they would get. Also people can declare interferences that slow you down and cost you. TL: Then won't it get so no one will ever disclose? GS: There is a difference between patents and applications. Applications are almost never disclosed. IEEE 2003 statement (150,000 members): until issued, scope of claims is not determined, so there is no value to disclosure of applications. DM: Do they understand members may alter claims based on SSO participation? GS: They understand two things. One is that the description is the invention. Rambus did not take Jedec technology. It was their own. And two, Rambus had NDA's with many of these companys and they saw others using their technology and amended their claims accordingly. DM: That's the case regardless of where you got the information? GS: There is nothing different about a SSO. Kelly said you can take it back to your company and be sure patents are written as well as possible. PH: Payne: spoliation? GS: Rambus has explained that at the time of becoming public they implemented a document retention policy. It stands uncontradicted at the FTC that Rambus consulted with a lawyer to implement this policy. The purpose was to save documents that might be involved in litigation but also to save the money of paying attorneys to sort through loads of extraneous documents. And the ALJ has stated that there is no showing of any relevant document being spoliated. TL: A silly finding on his part because really how can anyone be sure that something of importance wasn't destroyed? GS: Someone else could have produced a document that Rambus did not. They didn't. Or you could look to Rambus meeting minutes to for corroboration that all the appropriate documents were retained. PH: Rambus did not disclose anything about evaluating claims, altering technology, reverse engineering, or infringement analysis? GS: They maintained it all. DM: How do you know? GS: PH: Ten thousand pounds of documents were destroyed. That's a lot of documents. GS: Actually, we've looked at that. Let me put it in perspective. It's two very full trash cans per employee per month. PH: Counsel is familiar with the Sylvester holding in 4th Circuit Court? GS: Two lawyers very familiar with document retention, Dean Johnson and Diane Savitch advised Rambus. PH: Why did it extend to Rambus' outside law firm? GS: It is standard procedure after a patent issues to conform files with counsel. Why after 1998? Because they did nothing until the patent issued. PH: Rambus submitted a white paper to the FTC saying it was not seeking patents to SDRAM during membership. GS: I will not speak for that. That was a mistake. We were not then counsel. A subsequent white paper was submitted. PH: Was it the same lawyer advising on document retention? (laughter throughout the court room, doubtless some of it uncomfortable laughter) DM: You characterized as "refusal to deal." I can see why you want to shoehorn it into (I believe the reference was to case law). Gut if you don't want to deal, you don't go to the SSO. GS: Doug Melamed will answere more of this but let me say Rambus acted in good faith. They did not misrepresent. They did not push a feature or technology. When asked to comment, they declined. The refusal indicated to Jedec (per G Kelly) they were on notice. Rambus had legitimate reasons not to comment. There were two declines, so if someone thought more disclosure was appropriate, then they could have asked for it. But they didn't expect it. Even after 21i published--it was never adopted--Gordon Kelly said IBM would not disclose patents. TI, IBM, Micron with burst-EDO, Rhoden with SyncLink, the memo from Townsend on the members' manual was that if you presented you disclosed. DM: When Richard Crisp went back to Rambus, it seems it wasn't Rambus expectation. GS: Mr. Olover whowed part of Richard Crisp's Micron testimony. He didn't show the next page which said, "Q. What should you do? A. What IBM and others did." DM: But why the internal discussion? GS: They disclosed the '703 and then they discussed why or how they should continue and decided to only as required. JL: In complaint counsel's brief, pp. 20-21, IBM, etc, said policy was disclosure. GS: Contemporaneous behavior is inconsistent with that. JL: Was HP's counsel lying? GS: I wouldn't say, but their recollection is inconsistent with contemporaneous evidence. JL: Three patent applications were disclosed. GS: That's right, either by a presenter or a tattler. JL: There's a business advantage to not disclose. GS: (missed this) JL: (this, too) GS: Micron, Burst-EDO. JL: Back to Commissioner Leary's point. If you had the best technology why didn't you disclose? PH: And why the disclosure of patents on departure? GS: Dell was in the air at the time. It wasn't required but in trying to put issues to rest they did so, although it clearly did not accomplish that purpose. And one was inadvertently left off. PH: A critical patent was left off. GS: The '327 was never litigated although complaint counsel would have you believe it was. It was an implementation-specific patent, the type which was never discussed by anyone. It was circulated by someone else at Jedec. There is no indication that anyone ever looked at the letter Rambus sent or relied on it. End of first of Stone's presentation. I'll finish up with Melamed, Stone, and CC rebuttal tomorrow.