Rambus (RMBS): Trial notes 3/23/06 (2) Not providing...

Reply Private New

Replies (2) Next 10 Prev Next

Send PM Follow Ignore

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

calbiker

Re: None

Friday, 03/24/2006 7:56:02 PM

Friday, March 24, 2006 7:56:02 PM

Trial notes 3/23/06 (2)

Not providing bio of Teece. You can read some here.
http://www.haas.berkeley.edu/Faculty/teece.html

Rambus is lucky to have such a sharp and crafty witness. He left the entire court in stitches (including Whyte) when Furness was making a point that many companies were exiting the DRAM industry, including Intel. Furness: Isn’t it true that you can make a lot more money selling processors than selling DRAM? Teece: Yes, if you’re Intel. Talk about a stupid question! LOL

Whyte limited Teece’s testimony to only US sales. There was a second supplemental report written by Teece that contains world wide data. As I understand the pre-trial hearing, this report was too late to be entered as evidence. Teece only had one day to submit the report.

Stone questions Teece about its business model. Rambus focus on innovation. It does not manufacture but licenses its technology. Other companies manufacturer. Rambus doesn’t need patents from other companies. [Lets the jury know they really have no use for Hynix’s patents.]

Teece states the damage award should not be less than a reasonable royalty. He goes on that there are two types of royalties: lump sum (up front), and running royalty (% of sales).

Teece showed some slides of this calculations and assumptions:

Damages = $Base * % Rate

$Base is the US sales. Not what’s make here and exported.
%Rate is the royalty

Teece got sales data from Hynix. He divided the DRAM into two groups: SDRAM and DDR & DDR2. Start of infringement is June 23, 2000. The sales data ended Dec 2005. The total US sales totaled $4.4 Billion.

When Stone asked for the world wide sales, Furniss objected. There was a sidebar. Sidebar got resolved and Stone asked question again. Teece replied $18 Billion. Whyte then spoke to the jury. “Rambus is not entitled to ww sales, but you can take it into account.”

To make it easy for the jury, the ratio 4:1 is spelled out.

Teece then explains he used the rates 0.75% for SDRAM and 3.5% for DDR & DDR2. He obtained those rates using Georgia Pacific case law. There are 15 factors. The main factor is #15. What would be agreed to in a hypothetical negotiation conducted in 2000.

Other assumptions are: patents are valid, assume product infringes and they come to an agreement.

Teece got the rates by viewing 7 companies that singed license agreements in 2000. All were at the same rate (0.75% and 3.5%) except Hitachi. They were at 1% and a little higher on DDR. Rambus showing if a company litigates, they will pay more.

Teece explains the difference between a hypothetical and real license.
1. validity (in real license patents are untested)
2. infringement takes place (should pay more)
3. no up front payment (Teece is conservative, not inflated in R. favor)
4. need higher royalty base for US sales. Others are ww.
5. only for patents sued, not all patents (shouldn’t be any difference)

He says there’s no adjustment for raising rates because the uncertainty of validity is removed. When the companies signed in 2000 the patents were untested in court. He called it an uncertainty discount. Once infringement has been established, companies should pay more. The first 5 differences are reasons to increase the royalty rate, while the last difference listed provides reason to lower rate. Despite the differences, Teece stayed conservative and applied same rate as was used in 2000.

He then provided a survey of reasonable royalty rates (Lay Nuvell??).
1. Minor improvements: 1 – 5%
2. Major improvements: 3 – 8%
3. Revolutionary: 5 – 10%

The rates should be lowered a little because pharmaceutical rates tend to be higher.

A second survey provided a breakdown of the semiconductor industry (Gorden Smith & Russel Parks ?):

Out of 78 licensed, mean = 4.6%, median = 3.2%

SDRAM is much lower, while DDR rate is in the middle.

Teece’s end result for damages:

SDRAM = $12.7M
DDR = $95.9M
Total = $108.6M

He concludes calculation is conservative.

One of the Hynix pre trial motions was to prevent Teece from stating his calculation is conservative. Looks like Hynix lost that motion, as Teece used ‘conservative’ over and over to explain his calculations. Ha.

BTW, the mailman just delivered Cal’s jury summons. Ha.
Perhaps it's Phase 3 in SJ??? We'll have to check it out.

One more section to follow.