TMF's resident patent counsel (alleged) and RMBS skeptic begins to reconsider the legal odds for the bus in Virginia:
leviticus.aolboards.fool.com
Hi folks,
Just wanted to pass on that I have now read both briefs, by Rambus and IFX, on the issue of the applications that should have been disclosed to JEDEC, and the “first showing” principle. Let me take these issues in reverse order.
DDR
(1) I think IFX should lose on the “first showing” issue as to DDR SDRAM. It strikes me that the testimony of the JEDEC rep (McGhee) was more pertinent and more concrete, and he defined the duty to disclose as being triggered when the ballot for the standard is circulated (or words to that effect). The notion that one must “react” to every proposal made or lose you rights is absurd; you'd never be able to comply with a requirement like that – its too unworkable and not commercially reasonable.
(2) This leads to the next natural question of: does the duty to disclose nonetheless extend past the membership period as to “contemplated” or later filings so that IFX can win on this anyway? Remarkably, this issue is never articulated anywhere in the briefs, as far as I can see, in this fashion. My take on this is: it depends. A) IF you later filed applications using information from the group that you learned while you were there, then I believe yes, you are still bound by that duty because it “tags along” with how you derived the information in the first place. B) IF you later filed applications solely based on public domain information, then there IS no duty associated with such.
(3) So which scenario do we have, A or B? For the patents that Rambus sued IFX on, Rambus mentions (on page 15 – again, not great advocacy – why wait this long to tell your side of the story?) that their attorney did this work in November 1998, and that he did so without looking at JEDEC materials. I think this latter point is completely non-defendable based on the fact that he had exposure to such materials before this time during meetings with his own personnel, some of whom were at JEDEC. I'd love to see his notes on that, and what he had in front of him when the “idea” came to him suddenly to start claiming a programmable latency register in the late year 1998. Furthermore, at least one of the patents (5954804) was filed in February 1997, so I think he is grossly exaggerating on that one.
(4) In any event, he should NOT have done this work on these applications; he should have clean-roomed it out to someone else to avoid any taint whatsoever. My impression is that by that time, there was plenty of public information on the DDR and SDRAM standards, so he could have had someone else do it who was not contaminated. Nonetheless, even without this clean path he should have supplemented his testimony to the effect of “and by the way, I only needed to read IFX's own public data sheets on their parts - not information from JEDEC - to craft the claims for the patents in suit”
(5) Thus the real interesting bombshell is this, and again I can't believe it isn't articulated better in the briefs: It IS true that Rambus filed patent applications WHILE they were in JEDEC that they did NOT disclose. However, NONE of the patents that Rambus sued IFX on are part of this group! Rambus was quite clever to make sure they did NOT pick any of the set that might have some … unhelpful baggage.
(6) The IFX papers are amusing on this, because you can tell they understand this point quite well, but try to gloss over it by incorporating the earlier “proposals” into the mix, which proposals were made in the early 90s, and by talking about earlier applications that aren't even in the case. But the fact remains that none of the patents they were actually sued on were in existence during the times they complain about.
(7) Accordingly, if I were Rambus, I don't know why they did not accentuate this fact right up front, and say “Hey Judge, those case we didn't disclose, guess what - who cares, we did not sue anyone on them. We're happy to license THOSE on any terms you like. Also, there is no cause of action yet, because we did not try to enforce them. Furthermore, there can't be any harm unless I sue you or threaten you on them first So as to THOSE cases, how can there be any fraud?” I seriously don't think the mere act of not disclosing is what is actionable, or everyone would be in trouble. There is simply too much to keep up with and these companies have massive patent portfolios. What IS actionable is trying to ENFORCE it later.
(8) As to the patents in suit, they were all “filed” after Rambus left, as discussed above. The difficulty now revolves around point (2) above; were they still “bound” to disclose after they left? At this juncture, IFX did a good job of showing a funneling of info from JEDEC to Rambus, but, shockingly enough, only for patents that Rambus did not sue them on. IFX's ONLY point on this (the actual four patents) is on page 13 of their brief, point #7. This, in my mind, would not be enough to sustain a fraud charge. The fact that the patents in suit are “offspring” of the earlier applications is not enough of a tangible connection; the fact that they cover similar “technologies” to the earlier patent applications is also irrelevant in my mind. If IFX had some smoking gun such as an entirely completed draft from 1996 by Rambus that they simply held onto until shortly after they left, then I would be more convinced.
Vanilla SDRAM
(9) There WAS a first showing for vanilla SDRAM while Rambus was at JEDEC, so I think IFX would win on this point. But I'm not sure it matters for several reasons.
(10) The question then becomes whether Rambus' duty was actually triggered because they had applications on file that covered the standard. As you can see above, it looks like there WERE cases, but, these are NOT being enforced against IFX. So, there can't be any harm, as far as I can tell. The patents come from cases that were done much later, so they can't be subject to the same duty (unless they were derived from tainted information as discussed above).
(11) Let me digress for a minute before answering this. From what I can tell, privately, Rambus HAS been trying to have their cake and eat it, too, on this point. For some time, they HAVE told people that their patents cover the JEDEC parts. Now, realizing that this is a dangerous contention, they are backing off of this, because they realize it has serious consequences. But I don't know if these private communications are going to constitute an enforcement that is fraudulent under the law until they go to the mat against an infringer with such.
(12) For later filed applications covering vanilla SDRAM, see the discussion above.
(13) With that said, please understand the Judge may decide on a bright line rule. He may say, membership in JEDEC during a period in which you materially participated in the standard (vanilla SDRAM) entails an ongoing responsibility and accountability. You are stuck by your election, and the “duty” to disclose accompanies you forward. Otherwise, it leaves the door open to too much gamesmanship, and exactly the kind of problem we have now. I think this has an element of fairness that may appeal to him, and allows him to split the baby so to speak and give everyone a form of victory.
The upshot of the above is this:
* I don't know why Rambus did not draw the distinction between the patents in suit, and the (now irrelevant) earlier patent applications filed during JEDEC years. They could just as easily “jettison” these earlier applications by saying, “ok, lets make them a non-issue, we agree they are subject to JEDEC licensing terms. Now what?”
*I don't know why they did not very CLEARLY show why the patents in suit are completely clean, and derived solely from public domain information. The testimony from Steinberg should have been front and center, instead of burying it in the middle of the brief. My suspicion is that they cannot make this claim entirely in good faith, and that is why it is not expressed well.
At a minimum, they need to better attack IFX's only evidence on this point, consisting of limited circumstantial evidence presented on page 13 in two paragraphs. In my opinion, there is NOT enough there to sustain a fraud as to the patents in suit. Please appreciate that if Rambus were trying to enforce one of the earlier applications, my take on this would be completely different.
All in all, I think their chances are better than I posted the other day; probably 90% for DDR and improving to 50/50 for SDRAM. It all depends on whether the Judge feels that a membership in JEDEC is a “membership for life” for those proposals you voted on.
But don't trade this stock based on my opinion, please. This is just my analysis based on the briefs and the materials I have seen there, and I have been known to make mistakes (as my critics are so fond of pointing out…)
FWIW.
Da Greek
PS: I gotta go back to work, so don't bug me for a response for a few days... |
