Rambus (RMBS): Trial notes 3/22/06 There was a 1:20...

Reply Private New

Replies (4) Next 10 Prev Next

Send PM Follow Ignore

Followers	8
Posts	1020
Boards Moderated	0
Alias Born	08/31/2003

calbiker

Re: None

Wednesday, 03/22/2006 8:40:41 PM

Wednesday, March 22, 2006 8:40:41 PM

Trial notes 3/22/06

There was a 1:20 delay of trial due to pretrial motions.

Stone continued with JJ Lee.

I didn’t understand the first issue. Stone states: Hynix contends they do not use the rising and falling edge of clock. (Never heard this before) So Stone goes through the SDRAM datasheet where it states the edge of the clock is used.

Stone then builds a DRAM timeline of data bandwidth with Lee.

 
 EDO        ’91-’95   66-100 Mhz 
 SDRAM      ‘95-’98   200-400 bps 
 DDR        ‘99-’05   400-800 bps 
 DDR2        today    400-800 bps

Note, Mhz = bps (but those were the units provided)

Stone then builds a chart with Lee’s help of DRAM containing Rambus IP:

 
                   Sync   CAS Lat    Burst Length    Dual Edge Clock    DLL       Prog Write Lat   
 EDO              No       No                No      No                   No       No 
 SDRAM           Yes      Yes               Yes      No                   No       No 
 DDR             Yes      Yes                Yes     Yes                  Yes      No 
 DDR2            Yes      Yes               Yes      Yes                  Yes      Yes

Nissly got pissed that DDR2 was included. Stone’s reply was he’s making the point when Hynix wanted more DRAM speed, they took more Rambus inventions. Of course the jury gets comment from Stone again. Whyte allowed DDR2 in the chart.

Lee did comment that these elements alone do not make up the higher speed.

Stone went through some other issues which I feel not significant to list.

Stone was finished when high drama filled the room. Lee wanted to explain some of the testimony from yesterday. I certainly thought this is going bad fast. Stone asks Lee if he talked to any lawyers. Lee: No. They decided to let Lee speak his mind on Hynix clock. Ha.

Brown conducted the cross. He didn’t have much to say. He provided a correction to when the first DDR was produced. No big deal. Then I believe Brown gets into datasheet timing charts. I’ll tell you, I was already saturated with technical stuff, and I could see the jury wanted to strangle Brown. They were a sad sack, totally tuned out. Then Lee starts with his corrections from the day before. Nobody wanted to hear anything more about timing diagrams. But he goes on to explain details of illegal data states. I don’t recall that being an issue yesterday, but that’s what he was concerned about. Guess it could have been a lot worse. Ha.

Dong Chung came next. He’s VP of IP. He’s Rambus’ main contact during contract negotiations in 2000. Even though Chung is quiet and slow to speak, the jury quickly awoke, with many on the edge of their seats (at times). Stone was good. He made a convincing argument.

Negotiations were conducted from June 2000 to Sept 2000 when Hynix unexpectedly filed a lawsuit. Stone went through all the memos and emails during this time period. Lee kept repeating (in trial) over and over that the royalty rates were extremely high. That’s why they never signed. Rambus was asking for a 5 year term $25 Mil up front, 1% for SDRAM and 3.5% for DDR. In all the correspondence, Stone shows Hynix never protested to high royalty rates.

On August 9, 2000, Steinberg sends Chung the official agreement. On Aug 19 Chung replies with comments on agreement. He states that he’ll only discuss important issues in this email. His immediate concerns are:

1. R might offer lower rates to other companies. (Most favorite nation clause)
2. That everybody in the industry will sign-up within one year.
3. Tax issues, definition of product and a few other things.
4. Secondary concerns can be covered later.

Again Stone shows there’s no mention of the rates being too high. Chung counters that royalty rates were non-negotiable (according to Rambus). Stone states that Hynix knew the $25 Mil up front payment was negotiable. Yet not mention of that either.

On Aug 22 Steinfeld replies to concerns. The rate guarantee is declined. A comment was made that if Hynix doesn’t take the offer now, the rates will only go up. [this may have helped sending them over the edge]

Meanwhile, Sept 6 meeting was set for Tate and Park to finalize contract.

Aug 26 Chung writes another memo to Steinfeld that’s never sent. Chung requests that royalties should be based on memory chip net sales, not module net sales. And another issue on royalties on unfinished products. Again Stone mentions that Hynix requests no reduction in royalties.

Three days later, Aug 29 Hynix sues Rambus in this court.

Aug 30, Park sends Tate an email stating the offer is declined and that the patents are not enforceable and not valid. There was no word that Hynix sued Rambus.

Stone comments:

Since Sept 2000, Hynix had $4.5 Billion sales of DRAM in US and $18 Billion world wide.

Hynix must know that patents expire in 4 years.

Not quite sure on last question to Chung. Chung didn’t understand the Rambus inventions, or the elements contained in the invention. Stone replied “You’re the VP of Intellectual Property‘.

That was about it.