Rambus

[Aussie FX]

I've been hearing a lot lately about Rambus taking companies to court with Nvidia being the latest target.
http://arstechnica.com/business/news/20 ... rambus.ars

Then Samsung settled out of court recently for $900 million.
http://www.tweaktown.com/news/14056/900 ... index.html

What is all this about and what patents have allegedly been broken?
A lot of people seem to be calling Rambus the patent troll.
Sounds to me like something underhanded is going on.

[M_S]

Rambus has some 700 plus patents on DRAM technology and related implementations. There are several patents on the DLL design inside the memory chips, which I assume is one of the things they might have settled with Samsung, aside from the obvious programmable CAS latency issues that went in and out of court for years and years. Then there are a number of patents on system level implementations of memory controller adaptations to changing bus topologies, like for example the change in memory drive strength depending on how much load is on the bus, for example if the same controller is used for a 512 MB card and a 1GB card.

Intel and Rambus have cross licensing agreements in place but I don't know the situation for AMD, though I am fairly certain that ATI did not have anything.

Rambus is a very interesting company. On the one side, they have a very good group of inventors and on the other side a very good IP group. If everybody else is sleeping then it is not Rambus fault but that is of course not the complete story. It's a little bit like everybody bitching about Angelina Jolie and how she stole Brad Pitt.

Of course, they are also going overboard with some of the things they are doing but so did everybody else when it came to fighting them. I remember Micron's legal department leaking stuff out to me for writing a biased article on Rambus (of couse, they did not tell me it was biased), Infineon and Micron tag-teaming to sack the EMS patent portfolio by driving the company in the ground ..


And now one company is successful, and of course you know what that means. In highshool girl parlance, she is the slut, just because she is better looking. Now wait until somebody leaks the sex tapes, even if they were by themselves.

[Aussie FX]

Good ole jealousy huh?

I've heard that AMD are paying royalties, so they should be ok.
It's interesting aobut ATi though. After Nvidia being dragged through the courts one would hope they do the sensible thing. At least they have a CEO who's not an egomaniac.

Patents can cerainly get peoples blood boiling especially those on the side who have to pay the licensing fees.
There was a fair bit in the media here not long ago about royalties not being paid to the CSIRO for wifi.The CSIRO is funded by the Australian taxpayer as a research organisation for pretty much anything and they invented wifi. Unfortunately not many companies are paying the licensing fees with intel being the worst of the non payers so the CSIRO currently have lawsuits flying left, right and centre. I suppose they're waiting for the video to surface as well.

[M_S]

Litigation is the "Sport of Kings", it is incredibly multilayered and any agression can immediately kill you if you don't watch out. I have been involved in many of those things and as a rule of thumb, the one thing that materializes is that if it is true intellectual property that is really valuable for the general public, then the patent holder generally prevails. If the patent is one of these bogus patents that are just assertable by the letter of the word without having really "taught" the public about the new technology, then the patent goes under most of the time. There are new regulations in place where hearings are done to make sure that not just the law is served but also that the law is applied to its purpose and not just in a "random shot-gun" approach, i.e. let's just patent everything and see whether anybody in the future uses something related to that.

In the US, the process is called "Markman Hearing", named after Markman vs, Westview instruments http://en.wikipedia.org/wiki/Markman_hearing. In short, a jury hears both parties which then present their point of view as to what aspects of what are really covered by the patent, taking into account the body of the patent, the patent landscape at the time of the patent filing and the general practice of technology including "obviousness" to decide what it really was that the patent brought to the table with regards to novelty of the invention. Then a judge decides within the limitations of the patent law on the exact "allowable" interpretation of the claim language.

Needless to say that there is always the possibility for bogus "Expert Witnesses" and whatnot but both parties have the same chances and it becomes less of a matter of financial burden than of actual knowledge of the subject matter.

All of this requires the knowledge of the law and the technology, you can have all the facts right and you present them in a "wrong manner" and you lose. You have to anticipate the opponent's strategy and you have to have good data, then things will usually work out in your favor.

[Aussie FX]

Sounds like utter madness to me, Ive never really looked at patents before but what you're explaining seems like nightmare on elm street if you want to design/invent something.

So when you designed the NIA you must have been constantly checking to see if you step on anyones toes.
Did you have to get lawyers involved at any stage to double check things?

[M_S]

I agree, the patent landscape is somewhat confusing but with a bit of homework it is still possible to navigate through this. Keep in mind that the purpose of the patent system is still to teach the world a new invention while protecting your rights to it for a certain period of time.

In other words, you are trying to protect yourself against somebody just blatantly ripping you off by simply copying your stuff. Of course, there are nuances in the levels of ripping off and that is where things are getting complicated. Rule of thumb is that the broader a patent is, the wider it can be applied to those "rip-offs" or infringers or products that "have a substantial technical similarity with the claim language" but at the same time, the patent becomes very vulnerable with respect to somebody else having done already what you claim.

The commonly used example is the parabel of the chair.

Let's say somebody invents a chair, which is essentially an apparatus with a surface to sit on and a support structure between the surface and the ground that it stands on. Well, this applies to even a rock but the rock anticipates the patent, so that patent could be invalidated very easily.

So now you invent a chair with four legs. That patent does not apply to rocks and logs and whatever and clearly defines a specific construction. It is not applicable to a tripod or even a footstool (with a single foot) so none of those can be used to challenge the patent either. But, here is the kicker: somebody had a chair with 5 legs before you filed the patent and if you were to be bold enough to try and sue any manufacturer of a 5-legged chair because it has 4 legs as claimed in the patent (plus 1 extra), then these guys would be quick to point out that they were first.

The bottomline is that if you really invent something instead of copying things from others, you have a reasonable chance of your patent being upheld no matter what.

In the case of the nia, we licensed the core technology from Brainfingers who have several very old patents on some of the fundamentals and then we surrounded the existing patents with the new ideas that came up during the product development cycle. These new patent aplications are still in the "pre-publication stage, so I can't talk much about them but they are very novel so I am not overly concerned.

Also, to answer your question, yes we are working with good patent attorneys but I have myself also worked for several years in intellectual property management, so I have a good feeling for the field.