Comments on Rambus Column

I got a some hate mail over last Sunday’s column on the FTC’s suit against Rambus and Unocal. I’m posting some of the letters on this page, and plan to add more later.

Some people didn’t leave their names, others asked for anonymity and some sent identical letters, suggesting an orchestrated campaign is at work somewhere. Note: I will not post letters containing gratituitous personal attacks.

Rambus did nothing that Microsoft, Intel, and AMD haven’t done, are doing, and will continue to do. I could go on about GM, Ford, BMW, Toyota, ahh, I run out of typing time. These companies aren’t the problem, nor is the FTC. The problem is lawyers. To damned many of them. I expect, any time now, a lawsuit against Dell, (or any other PC maker), because a serial killer used the Internet to track down a victim. Take it to a logical conclusion, let us sue Kingston or Rambus for making the RAM, and Intel for making the processor. Not Microsoft, we already have seen how good their lawyers are.

see, this is where you need to do some investigating. Our u/s is that what was used by the MMs in the SDRAM spec were critical features from the RMBS RDRAM design. Features that the MMs knew about thru the NDA’s RMBS presented to the MMs in trying to get RMBS to adopt their WHOLE design.

instead, it appears the MMs cherry picked the critical features from RDRAM needed to make SDRAM work. So with that as a starting point, WHO is wrong here?

last, under our patent laws, RMBS has the right to amend their original patent to cover ONLY what was presented in the original patent. That according to the CAFC allowed RMBS to do what it did LEGALLY.

now whats interesting in your articles, you don’t challenge the MMs actions regarding the situation they forced RMBS into. Maybe instead of just pushing the FTCs allegations into the media and then justifying them, you could do a little digging into just WHY RMBS did what it did. WHY don’t you ask the MMs to show us what alternate solution they would have used to make SDRAM? But something that was all THEIR idea, not using cherry picked designs..

Think thats what they call that fair and balanced reporting…

I think you are still missing the story of Rambus’ IP and the key events
in this matter. Rambus invented the IP in question used in the SDRAM
standard and JEDEC members learned of these Rambus inventions through
NDA’s they had with Rambus long before Rambus joined JEDEC. Under NDA,
Rambus taught the DRAM industry this very valuable technology, and the
memory manufactures incorporated it into the SDRAM standard developed by
JEDEC. Rambus was neither a presenter at JEDEC nor did they ever vote
on any measure while a member of JEDEC; they met their disclosure duty
to JEDEC as judged by the Court of Appeals for the Federal Circuit (see
their 1/29/03 ruling). Rambus “hoodwinked” no one. Rambus did not “end
up with key intellectual property;” they owned it from the get-go.
Rambus amended their patents to add claims to cover SDRAM. This is
totally legal and is even encouraged by the US Patent Office.
Seven of the 10 DRAM manufactures agree with the above and are paying
royalties to Rambus on SDRAM. The three who are not, Infineon, Hynix,
and Micron are desperate and are doing their best to evade paying
royalties. It’s as simple as that.

I would like for you to consider a different
viewpoint on Rambus’ ethics. Rambus initially joined
JEDEC with the intent of promoting RDRAM as the next
JEDEC memory standard. They were invited by one of the
DRAM manufacturers.

At this point in time, they had shared the RDRAM
inventions with all of the major DRAM vendors and had
non-disclosure agreements with all of them. All of
these companies were cognizant of Rambus technology
and were well aware of the fact that they were
incorporating some of this technology into the SDRAM
standard. Some of them expected to pay royalties to
Rambus and are currently doing so. Others (primarily
Micron and Infineon) believed that they could bully
Rambus in the courts into giving up their IP and are
currently trying to do so.

Rambus was faced with an ethical dilemma during their
tenure with JEDEC. They, and all of the DRAM
manufacturers were aware that the SDRAM spec would
utilize Rambus IP. Rambus had been asked by the USPTO
to divide their original patent into a number of
separate inventions and were in the process of doing
this. One of the US patent laws states that, if a
patent claim is common knowledge prior to being
issued, it is invalid so they could not reveal their
intent to file these claims or they would lose them.
In addition, they were, as you pointed out “swimming
with sharks” and some of these sharks (particularly
Micron) would have done everything in their power to
subvert the patent process and deny Rambus the rights
to their IP if Rambus had revealed their intentions.

So, what would you suggest that Rambus do in this
situation in order to be “ethical”. I don’t see how
they could have done anything much differently without
totally giving away their technology. The one thing
that they could have done was resign from JEDEC when
it became obvious that JEDEC was not going to adopt
RDRAM (about a year before their actual resignation)
but this would not have changed the situation in any
significant way.

Rambus, did not, however, just “end up” with IP
covering SDRAM as you stated in your article. The IP
in the Rambus patents was always theirs and was built
into the SDRAM specs by people who were well aware of
what they were doing. Rambus did not change their
inventions after the fact, they simply clarified them
at the request of the USPTO.

But you are so, so wrong here. JEDEC, like all standards bodies, is now
required to consider — and not reject out of hand — patented technology as
part of its standards. In this case, Rambus had developed a new type of
memory (first generation RDRAM) and was invited to join JEDEC by what we now
know are scheming DRAM manufacturers. When Rambus realized the plan was not
to perhaps standardize on RDRAM or to incorporate features of it openly with
the allowance of a reasonable and nondiscriminatory royalty, but to quietly
rip off those features giving Rambus no credit of royalties, they tried to
protec themselves. Obviously it was a snake pit of interests, but the FTC
action cannot show any harm to consumers, and comes down on the player least
guilty of misleading actions, not ot mention the smallest and
least-sophisticated at the time.

Amusingly the FTC has accused other standards bodies of anitcompetitive
behavior when they refuse to include patented technology as part of a
standard — in one case, keeping an innovative toilet valve out of a
standard to protect manufacturers of toilet valves from competition.


This entry was posted in Archives. Bookmark the permalink.