Rambus (RMBS)

[Skeptic]

Hope Stk does not mind the repost here from TMF, but it further adds to the debate...

And then there is this from Calbiker on ihub;

“Speaking of which, look at all the recs for stk_hawk's CAS latency post. It's wrong. Doesn't anybody do their own thinking there?”
http://www.investorshub.com/boards/read_msg.asp?message_id=10003735

“About 4 weeks ago someone listed the claims going to trial. If that list is correct, then stk_hawk's list is incorrect. It's extremely dangerous to conduct a patent search using 'rambus' and the last 3 digits of the patent. Some Rambus patents have the same last 3 digits. Looks like he has the wrong '916 patent listed. His conclusions are as well quit iffy.”
http://www.investorshub.com/boards/read_msg.asp?message_id=10004090


Well with all due respect (and I do respect Calbikers technical analysis – i.e. difficulties in DLL design around for example) his infringement feature analysis is incomplete because it does not comprehend the issue of dependent claims.

But first of all, just because I only listed the patent abbreviation in my original post (i.e. '918 and '916) that does not mean I did not have the full patent numbers or that I simply did a patent search using “rambus” and the last 3 digits of the patent. I did my analysis using the full patent numbers as listed in the legal brief which identifies the 10 claims Rambus selected.
http://investor.rambus.com/downloads/2005-01-21%20Rambus%20Election%20of%2010%20Claims%20for%20Trial...

So with that out of the way, let's looks at the dependent claim issue.

A closer look at the 10 claims Rambus chose to take to trial, reveals that all of the claims are dependant on other claims. For example, claim 34 of the '105 patent is also dependant on claim 31. What does this mean to Greg Stone? Well it means that in order for Greg Stone to convince a jury there is infringement of claim 34 of the '105 patent, he also has to prove infringement of claim 31. As a consequence, there are other Rambus IP features contained in the dependant claims, which Calbiker's analysis fails to comprehend.

Here is a summary of the dependant claim issue for the 10 claims Rambus selected for trial.

'105 claim 34 (dependant on claim 31)
'918 claim 24 (dependant on claim 18)
'918 claim 33 (dependant on claim 18)
'120 claim 33 (dependant on claim 29 which is also dependant on claim 26)
'020 claim 32 (dependant on claim 31 which is also dependant on claim 30)
'020 claim 36 (dependant on claim 35 which is also dependant on claim 30)
'916 claim 9 (dependant on claim 1)
'916 claim 28 (dependant on claim 26)
'916 claim 40 (dependant on claim 26)
'863 claim 16 (dependant on claim 15 which is also dependant on claim 14)

As a side note, this dependant claim issue also has ramifications on Hynix's invalidity arguments at trial (i.e prior art defense). For example, in order for Hynix to invalidate claim 33 of the '120 patent due to prior art, Hynix must produce prior art that not only invalidates ALL of the elements contained in claim 33, but also ALL of the elements of claim 29 and ALL of the elements of claim 26.

It would appear that Calbiker does not comprehend this requirement of patent law either, but I find it very interesting that the resident patent lawyer from the “firm” isn't educating Calbiker on this aspect of patent law either. Perhaps it might complicate his simple DLL design around arguments he is putting forth. Imagine that, other issues that need to be designed around other than just DLL.