Rambus Antitrust Trial Begins

[drscott]

For a long time most of the public who knows anything about it has bought the propaganda (from DRAM manufacturers colluding with AMD) that Rambus went to JEDEC meetings, secretly patented what they saw there, then tried to profit by suing makers of JEDEC-standard parts. Completely false, and shown to be so in court case after court case. So the press should have fun with what's about to come out in the antitrust suit, Rambus vs. Micron, Hynix, etc.

Jury selection began today, trial actually starts in 2-3 weeks, depending on how quickly they get a jury for what promises to be a 2-3 month case. Since the defendants (Micron, Hynix, etc.) have already plead guilty to pricefixing DDR, which was part of the scheme to defeat Rambus memory and steal its component innovations, it will be hard for the defendants. Here's a bit of transcript from arguments in an earlier case which gives you the flavor of it:

He was concerned that if Intel implements Rambus, all other applications will move in that same direction, and even if the architecture changes, their signaling, the Rambus signaling, will remain for many generations

Stone


1 Let's go back to that document 777 we looked at

2 earlier. I only showed you the part about wanting to

3 get the same performance as Rambus. Now we look at the

4 part that precedes it. This is what Farhad Tabrizi was

5 going to say to everybody when he sent out this email.

6 This is what gives us our first clue as to what was

7 motivating the DRAM manufacturers at the time. He's

8 really worried, because if Intel and Rambus get

9 together, put together this marriage of the Rambus

10 technology and the fast Intel processors, we saw IBM

11 years earlier thinking they could corner the PC market,

12 but what's he worried about?

13 He's worried -- and it's in the second

14 paragraph that I've highlighted -- that we will become

15 a foundry for all Intel activities, and that if Intel

16 wants to do business with us, we may get a small share

17 of their demand or not. He says Rambus licensing is

18 not just an issue of paying a royalty to Rambus. It's

19 not. It's not just an issue of paying a royalty to

20 Rambus. He was concerned that if Intel implements

21 Rambus, all other applications will move in that same

22 direction, and even if the architecture changes, their

23 signaling, the Rambus signaling, will remain for many

24 generations.

25 So, they're worried. They're not worried about



1 paying royalties to Rambus. They are worried that they

2 have gotten so far behind in the race for technology,

3 they have gotten so far behind in their ability to

4 deliver faster speeds to the computer manufacturers and

5 faster speeds to the consumers, and they are afraid

6 that they can't keep up without using the Rambus

7 inventions.

8 So, what does he say? He says, "I urge you to

9 educate others and get their agreement to say no to

10 Rambus." Say no to Rambus. And then he goes on to

11 say, "What I showed you earlier, that SyncLink is the

12 way to go. We are going to all get together. We are

13 going to put together this SyncLink proposal. We know

14 it infringes Rambus patents, or at least Rambus tells

15 us it does. We think they're invalid, Rambus tells us

16 they're not. It doesn't matter. We don't want to be a

17 foundry for Intel. We're going full speed ahead, full

18 speed ahead on SyncLink." Full speed ahead using

19 technology that they know Rambus says infringes.

20 RX-1444. Samsung has announced that they're

21 going to start building an RDRAM product that Intel

22 needs. Samsung has made this announcement. A press

23 article reporting on that announcement is included

24 here. This is a Micron document, and bring me up the

25 second part of that. What does the Micron author write



1 when they see that article?

2 "This article shows that Samsung has broken

3 ranks with the other suppliers and sold their soul to

4 the devil." In other words, Samsung has decided we're

5 not going to try to go slow. We're going to get on the

6 fast pace of revolution, and we're going to get a

7 product out there that will go fast, but that,

8 according to Micron, was selling their soul to the

9 devil and breaking ranks with other suppliers.

10 And then if we go up to the top, what do we

11 see? An officer of Micron writes this email. He says,

12 "I've certainly made the point with the officers," and

13 you'll see in context he's talking about the officers

14 of Samsung, "that Intel is essentially disabling our

15 marketing, applications and design, and other key parts

16 of the company, and ultimately could control the DRAM

17 industry the same as they have others. I don't think

18 everyone considers it as much of a threat as I do.

19 There are a number of options for Intel. Seems to me

20 they'll be forced to do several strategies

21 simultaneously to avoid egg-on-face and the Justice

22 Department. 1, provide an alternative chipset; 2, use

23 Samsung to drive Rambus; and 3, work with the industry

24 on a non-Rambus packetized DRAM."

25 So, what was driving the DRAM manufacturers? I



1 offer this evidence not to show you that the DRAM

2 manufacturers had engaged in an unlawful joint boycott

3 of Rambus products. I offer this evidence to explain

4 to you why the DRAM manufacturers, with all this notice

5 about Rambus intellectual property, would decide we are

6 going to take the risk of using that intellectual

7 property because if we don't and if the RDRAM product

8 wins out, well, we're going to be a foundry for Intel.

9 Their fear of being a foundry for Intel drove them to

10 take the risk of using Mike Farmwald and Mark

11 Horowitz's inventions piece by piece and bit by bit in

12 the products that they designed at JEDEC, because they

13 realized the continual evolutionary, go slow, hay wagon

14 pace of JEDEC wasn't going to work.

15 So, where does that take us? It takes us to

16 the end of the evidence and the end of the three

17 issues. Duty? No duty. Breach? No breach. Any duty

18 that was enforceable under the antitrust laws? No.

19 Any exclusionary conduct? Has Rambus kept anybody out

20 of the business? The only extent to which Rambus has

21 kept anybody out of any business is by being the ones

22 to invent the solution to the memory bottleneck. And

23 they got a patent on it, a lot of patents on it,

24 because their invention -- and everyone concedes -- it

25 was revolutionary. So revolutionary that everybody is



1 using it today.

2 And did they do anything that caused any

3 anti-competitive harm? No. And I want to talk about

4 the anti-competitive harm for a moment. If Rambus had

5 never joined JEDEC, the world would be the same today,

6 except we wouldn't be here. If Rambus had disclosed

7 its IP at JEDEC meetings, we'd be where we are today in

8 terms of people using that technology, because they all

9 knew about the technology, and they used it anyway, and

10 the best example of their willingness to use the

11 technology despite being warned is SyncLink, because

12 Richard Crisp went to the SyncLink meeting, and he

13 said, you infringe our patents. And they said, no, we

14 don't. No, we don't. RamLink was first.

15 So, he went back to the next meeting, it was a

16 JEDEC meeting, and he said, RamLink was not first, we

17 were first. And he wrote them a letter and said, we

18 were first, forget that argument. You are not going to

19 invalidate our patents on RamLink.

20 So, what we know -- what we know is that the

21 but for world that we live in today is a but for world

22 where the DRAM manufacturers had a choice. Should I

23 sign up with RDRAM, pay royalties, make the fastest

24 possible product, give the consumers what they want and

25 need, or should we try to all hang together and go the



1 slow performance route? And there's a risk in taking

2 the slow performance route, because if you go slow and

3 some other manufacturer goes fast, you could get

4 knocked out of the market.

5 So, when Samsung decided they were going to go

6 fast and start making the RDRAM and break ranks with

7 the rest of the industry, well, the other guys going

8 the slow route got worried, and they said, we've got to

9 speed it up. And so they re-invigorated their whole

10 SyncLink plan.

11 And if we bring up 857, I think this is the

12 right one -- yeah, that's it. Bring up 857. Here's

13 what they did when they became aware of this. Bring up

14 the highlighted part, if you will.

15 They said, oh, my gosh, we've got to do

16 something about this. They called a SyncLink executive

17 meeting in Yokohama, Japan, and the purpose of the

18 meeting was to re-ignite the consortium in light of the

19 recent Intel announcement to enter into a design

20 partnership with Rambus for the so-called Rambus

21 Direct, and with few exceptions, every DRAM

22 manufacturer was represented there. So, they knew,

23 Intel is such a threat to us -- it's really Intel and

24 not Rambus, because Intel is a dominant force in the

25 market, and it is fair to say that Rambus is not a very



1 large company or a very dominant company -- but they

2 knew that if Intel paired with Rambus, their slow

3 products wouldn't make it.

4 So, they called a meeting and they said, we

5 know we infringe Rambus patents -- we know they told us

6 that -- but we're going forward anyway. We are going

7 to re-ignite the consortium. We are going to ignore

8 the red flags that Mark Kellogg and everybody else

9 talked about, we are going to ignore all that

10 information, and we are going to go forward.

11 And what does that tell us? That we would be

12 today exactly where we are if Rambus had sent a lengthy

13 letter from its lawyers to everyone at JEDEC saying, by

14 the way, we have some patents, we have some

15 applications, we expect to get more patents, we think

16 they're going to hold up someday, and we think what you

17 guys are doing may well infringe them, and what would

18 have happened is the DRAM manufacturers would have done

19 in that scenario exactly what they have done now, which

20 is to say, we don't -- we'll take our chances.

21 We don't think your patents are valid. We'll

22 take our chances. If we have to, we'll litigate with

23 you. We'll litigate until something freezes over. We

24 will litigate, and we will invalidate your patents. We

25 are going to take our chances. We will not become a



1 foundry for Intel. We live today in the same world we

2 would have lived in if Rambus had told the JEDEC

3 members everything that complaint counsel envisions

4 should have been told.

Comments

[fuzzybearcub.livejournal.com]

Thanks for the additional info.

[pklexton.livejournal.com]

Interesting. I keep intending to read more about this every time you post about it but never get around to it. From the inquiries I've made of other lawyers I know about the case, I get the distinct impression Rambus is not very well respected to put it mildly - at least amongst my non-scientific sample that admittedly may have some self-interest bias, and I am aware as you point out that they've been cleared of at least some past wrongdoing. Money tends to talk even if (like reputation) it really shouldn't and there's a lot of money lined up against them.

The little bit of antitrust law I learned in law school (and not well apparently judging from the grade I got) is much "reinterpreted" after the proliferation of Reagan/Bush I/Bush II judges. (When it goes in the other direction it's called judicial activism.) Antitrust defense used to be a huge profit center of the major business law firms. Today, not so much.

[dr-scott.livejournal.com]

Well, their image is contaminated by a period when the cartel and AMD paid PR firms to spread disinformation, and AMD fanboys and Slashdot geeks who come after every IP claim thoroughly trashed their reputation. When TI left the DRAM industry, they struggled for years to get paid for their patents, and the remaining industry did not want to give up the cozy "we crosslicense everything and never pay" club. That Intel had designated Rambus as its preferred future memory created fear the DRAM makers would lose control of their future design freedom.

This is a special case where the industry was too stupid in their collusion to avoid leaving a mountain of damning evidence of price fixing and conspiracy to steal the component innovations of Rambus memory while driving Rambus itself out of the market. The Justice Dept. already got them to plead guilty to pricefixing during this period, and the evidence shows the scheme was to hold down DDR prices to convince OEMs to deny Intel and Rambus the new memory architecture they had worked on; then after they had succeeded, to raise DDR prices. Many of the execs involved were convicted and jailed.

This is a private antitrust conspiracy lawsuit. California law will automatically treble jury-awarded damages, set at $4 billion by expert's report and now larger with the passage of time. Rambus' reputation and market opportunities were damaged by the conspiracy, and every person associated with Rambus, past and present, wants them to pay up.

I know the guys who founded Rambus. They are guilty of naivete and nothing more. Every one of the accusations against the company were exaggerations and outright falsehoods designed to benefit the accusers.

[pklexton.livejournal.com]

Here's to Rambus getting justice and you making a nice profit. I have no dog in this fight. :)

[voxbaryton.livejournal.com]

still around mister?