Trial Notes 4/12/06 Part 2
Furniss gave some background on Horowitz: BOD on Rambus, owns 1.4 million stock.
F: In the chart you prepared the other day, there was no mention of the word Bus.
H: No.
F: But in the patent it’s frequently used.
H: Yes
F: Do you know which claims are asserted?
H: No, just have a general sense which are at issue.
F: Reads the summary of invention from patent where the bus is discussed. Is that why the company is called ramBUS?
H: No. Mike started it. It was a play on words coming from the popular film at the time RAMBO! The entire court burst into laughter. LOL. [A good lawyer should know what the answer is before asking. Furniss screwed up.]
A little while later the court phone rings (rather loud). As the court lady picks it up, Whyte yells to her “Say NO DEAL!” The entire court burst into laughter again. Was that is clue to Hynix? Deal time is over???
Furniss wasn’t deterred. More on the narrow high-speed bus. He reads from a 1992 Rambus brochure:
“Rambus designed a revolutionary chip to chip bus…”
H: We were quite proud of that and achieved a high BW.
F: Other inventions increase speed?
H: Yes
F: Did block size info appear in packets?
H: Yes
F: He make a comment that Hynix dram don’t use packets. As if that makes a difference. And Horowitz commented to that effect.
F: DLL- Is the invention to put it on memory device?
H: Yes
F: Dual edge clocking- Have you seen the TI prior art scheme:
H: No
F: Does Redwine & Novac ring a bell?
H: No.
More prior art discussions.
H: At the time DRAM was simple. Wanted to remove functions. The root of the invention was to go against this and add complexity.
F: Have you heard of IRAM (Intel)?
H: No
F: IRAM increased the complexity of DRAM by moving functions from the controller to dram. [Furniss is making a good point.]
Furniss asks about claims. Horowitz replies he knows nothing about the claims. Buy it didn’t stop F from asking again, multiple times.
F: Did R file claims that were extensive and fundamental?
H: Yes. We thought they had broad applications. It became clear the 2nd half of 90’s when others were using our interface.
Furniss talks about Karp. Karp thought lots of stuff wasn’t covered by R patents. The claims are not good. Did you review claims files by Steinberg?
H: Don’t think so.
F: Did claims Steinberg write represent your inventions?
H: I’m not an expert in claim language.
F: Why did it take 8 years to get these claims?
H: Company was focused on inventing new technology. When other companies started using Rambus IP, R started new patents. Early on, protecting IP wasn’t a main focus. [note, somehow, using R IP didn‘t look good] Ha.
Getting back to the drawing Horowitz created the other day. Furniss notes that the dram 500 Mb/s BW H wrote was for RDRAM and 2.5 Gb/s was for later R devices. Not for Hynix parts.
H: Yes, but they’re getting up there.
F: The Performance Gap - has it been addressed?
H: Yes From a bandwidth perspective, not latency.
F: He pulls out the Hennacy book from yesterday and goes to the performance chart.
H: The performance of mem vs. cpu talks of access time.
F: Did Hennacy forget about BW?
H: He’s a smart man. He was making a point about access time.
F: Rambus 2005 Developer Forum Japan- Has presentation from Woo. Rambus is telling the industry that the gap is widening with conventional DRAM.
H: Speed is ambiguous. Can talk about latency or bandwidth.
There were a few other items with H, but that’s about it.
It’s Miller time and only half finished.
