Rambus (RMBS): Trial notes 3/23/06 (3) Furniss condu...

Reply Private New

Replies (6) Next 10 Prev Next

Send PM Follow Ignore

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

calbiker

Re: None

Saturday, 03/25/2006 12:17:16 AM

Saturday, March 25, 2006 12:17:16 AM

Trial notes 3/23/06 (3)

Furniss conducted the cross on Teece. He started attacking the 7 companies that settled with Rambus in 2000.

Toshiba: A Japanese company. Are they paying now? Toshiba sold this section of business to Micron.
Teece: The rate transfers to the new company.
Furniss: Micron doesn’t pay.
Teece: I can’t assume someone cheats on an agreement.

Hitachi: They got out of the DRAM business.

NEC: They got out of the DRAM business. Hitachi and NEC DRAM business formed Elpida.
Teece: The rate transferred to the new company.

Furniss make is sound that Elpida had no choice but to accept Rambus rates. The new company (Elpida) couldn’t get any funding with a lawsuit over their heads. Teece: I don’t see it as any problem. Furness provides evidence: Electronic Buyers News Sept 8, 2000, analyst comments that NEC & Hitachi wanted to avoid any legal action. Teece: You want to come in clean into a new company, but not uncommon for a new company to have lawsuits.

Furniss brings up Teece’s expert report of 2005. Where Teece states 53.4% of litigated patents are valid. Not sure what the relevance was in regards to Elpida discussion.

Oke (?): Have they ever paid royalties on SDRAM or DDR? They only build SRAM.

Samsung: They had a 5 year license. Now expired.
Teece: What they do in 2005 is limited relevance.
Furniss: Have you used the Book of Wisdom? - can look at intervening period.
Teece: We can only use Book of Wisdom when there’s inadequate evidence. Here we have 7 companies.

Mitsubishi: Out of DRAM business. To Elpeda.

Teece: Labels have changed but DRAM is still being made.
F: Companies are exiting DRAM market. Why? (This subject got Stone wired. We’ll hear from him when the jury leaves.)
T: The business is cyclical; highly profitable in some periods and not in others.
F: From 1995 to 2000 DRAM business lost large amounts of money. 5 years of heavy losses.
T: But DRAM usage expanded. Gross margins is at 50%. This is good.
F: But not net profit.
T: That depends how you run the company.
F: Wouldn’t a companies ability to pay affect rates?
T: Just because its not making money doesn’t mean they can infringe. United is in bankruptcy but they still pay of fuel.
F: From ’95 to ’03 Hynix mostly had losses. $9.2 B accumulative. (Not 100% sure on that)

T: In the minds of the 7 companies that settled, there was sufficient clarity to pay. (not sure of the question)

F: Questions same the royal rate.
T: The companies are worried about paying more than others.
F: What does it imply when they all pay 3.5%?
T: Rambus has lots of leverage.

F: Rambus is allowed foreign royalties if they have foreign patents. Does Rambus have valid patents in Europe?
T: Don’t know.

F: Asks about Total Royalty Burden. I believe he states Weinstein calculated 2 –4%.
T: Don’t know if it’s relevant. Haven’t seen his work.

F: Hynix license with TI is a tad under 2%. TI has thousands of patents.
T: The key is what performance they provide. Typically 1% of patent portfolio is valuable.

F: Lay Nuwell study contains pharmaceuticals. Need to lower rate. Medical products have higher rates. Software has higher rates. Rates are skewed too high.
T: Cross licenses causes rates being too low.

F: What’s the value of the patents?
T: They allow more speed.
F: Have you heard of the saying ‘Want more speed, use more inventions?’

Stone’s turn with Teece.

S: In 1990 DRAM speed was at 25 MHz. H&F inventions increased speed to 500 MHz. A 20x increase in speed. In 1995 Intel was concerned of performance gap between DRAM and processor speed.
T: Unplug the bottleneck and you create value.

Stone goes back to the DRAM speed chart he created with Lee’s help and shows the value of the patents.
T: Rambus inventions provided uncommon value.

S: Horowitz is a pioneer. Received IEEE praise. He provided a revolution in the industry.

Here he is:
http://www-vlsi.stanford.edu/~horowitz/

S: Discusses letter from Chung to TI. Shows Hynix paying TI 3% and $77M for 6 months. Did TI give Hynix a better rate because they were losing money?
T: No.

Stone stresses a level playing field is main concern for companies. Same royalties.
[At this point in time Furness must be burning up inside as he can’t mention Infineon. Ha.]

Stone talks about Hynix $9B loss.
T: Hynix has access to capital. Has no doubt they can pay the royalties.

Stone discusses that about 50% of patents cases go either way. When signing unproven patents there’s a 50% uncertainty. If a 50/50 chance led the 7 companies to pay 0.75 & 3.5%, how much would they pay if they knew they were valid?
T: 1.5% & 7% is applicable.

Furniss takes over again:
F: There’s still a memory bottleneck with a 3.6 GHz Pentium processor.
T: That just goes to show these patents have uncommon value because nothing better has been invented since 1990.

The jury was dismissed.

Immediately afterwards, Stone lets Whyte know the condition of the DRAM market back in 2000. One can tell Stone was agitated when Furniss went through the list of 7 companies that signed licenses stating they are now either out of business or stopped paying. Stone states those companies that survived because of price fixing.

Whyte: You really believe this needs getting into?
Stone: Yes

Note, I’m not providing exact quotes, but paraphrasing many discussions. I may even have some wrong. Ha.

That’s it for Cal’s week of reporting. Perhaps someone else can pick up the slack? Let me tell you, it was a hell of an experience witnessing this court case in person. In addition, it was a pleasure providing this news to others. I’ve seen plenty of other investors at the trial who should be more than capable of providing court info here at I-Hub. If you got something out of this, then it’s time to step up to the plate.