ISSUE #4: DO YOU REALLY NEED TWO CLOCKS?
The fourth area of the specification that Samsung challenged in the reexamination deals with the use of certain language in claims 5, 9 -12 and 41 – 49. The Examiner lumped these into a category of claims dealing with clock signals which Samsung argues must be two separate clocks because the specification only shows two external clock signals. Hence, according to Samsung the addition to the claims of an embodiment in which only a single clock is used must be tied to the later filing, because it is not in the original disclosure.
As an initial matter the grouping of claims is a bit odd since many of them don’t seem to relate to the external clock signal issue. But that’s another story.
The Examiner so far disagrees with Samsung and says there is plenty of support for language that only mentions an (single) external clock.
Samsung’s argument on this front is very hyper-technical and again tries to borrow from bits and pieces of the arguments made by experts, the claim constructions from the Hynix case, etc. In other words, they again require the Examiner to consider a bunch of extrinsic evidence to the specification. And, once again, the Examiner clearly does not seem to be paying much attention to this kind of evidence.
Here is the nub of Samsung’s argument:
“…Under the Patent Owner’s proposed construction of an external clock signal as “periodic signal from a source external to the device to providing timing information (see Reply at page 45) the clock signal generated by the clock generator is not an external clock signal because it can only be an external clock signal once it is received from the external source. Only the clocks that are actually received by the devices, CLOCK1 and CLOCK2, would meet this proposed construction as they are the clocks that provide timing information to the device. In other words, from the perspective of the claimed memory controller, there are two clock signals: CLOCK1 and CLOCK2.”
This comes from their comments, pages 18 – 19 filed May 15. I must say I read this passage several times and it still makes little sense to me. I think the key is the very last sentence, which kind of gives away the real weakness in the argument. For Samsung to succeed, the Examiner must look at the claim “…from the perspective of the claimed memory controller…” Needless to say, I think this argument is doomed to fail, because one doesn’t get to kill claims based on an isolated “perspective”; the claim element is either there, or not. The tree falling in the forest does physically make a sound, whether or not it is perceived by others.
Also, Rambus rightly points to FIG. 8A to explain that there is, indeed, just one “external” clock signal that is sent to the device; all that happens is that this same clock signal is tapped at different taps along the clock line:
This is taken from their Reply filed April 15 on page 9. In other words, the CLOCK1 and CLOCK2 language is effectively just a label, not an indication that there are two clocks. The blue box I added shows where the two taps are used for the device 51.
Once more the reason for this attack angle is plain to see: Samsung could kill a lot more Rambus patents if its logic were adopted.
But so far the Examiner agrees with Rambus. Once more what jumps out at you from reading his analysis is that is very short, to the point, and he is focusing solely on the specification, and never mentions anything about either party’s experts, court rulings, etc.:
"As shown in figures 8A and 8B of the ‘696 patent, a single clock signal is sent, but is characterized as either “early” or “late,” hence one rise and one fall pulse (two clock signals) that are derived from the single clock signal".
Based on what I’ve seen I am reasonably sure and expecting the Examiner to maintain his reasoning on this one. There just isn’t enough there in the Samsung response to convince me that are going to turn him around on this one.
At this point of the clock cycle I show Rambus winning 4 – 0 on the written description issues.
I travel on vacation for the next two weeks, so the next installment will not show up until the end of this month earliest.