Re: Rambus uppcomeance

From: Robert Harley (Robert.Harley@inria.fr)
Date: Sat Mar 17 2001 - 07:01:15 PST


Warning: opinion laced with "if"s and "maybe"s ahead.

Our resident Grlygrl201@aol.com wrote:
>>It seems that SDRAM doesn't actually infringe their patents due to
>>technical differences [...]
>Too early a call on that, Robert.

Yes, so let me rephrase: there is a significant likelihood that SDRAM
and DDR SDRAM will be found to not infringe most of the patent claims
that Rambus currently makes a big chunk of its earnings from.

>"[...] the impression adopted by many observers is that Rambus' bus is a
>"multiplex bus" and common SDRAM and DDR memory uses just a "bus". Without
>diving into the technical minutia of the two definitions, it is a fact that
>many versions of SDRAM and DDR still perform "multiplex bus" functions [...]"

Maybe he should dive into the technical details, but I wouldn't expect
an investor type to do so. His remark is bogus. It doesn't matter
that SDRAM modules implement the same functionality. When they don't
use the multiplex bus mentioned in the patent claims, they don't
infringe those claims. And it seems that Infineon, Micron, Hyundai
and others may be in this situation. Disclaimer: IANJP (I Am Not
Judge Payne) etc.

>"For those not keeping score at home, there are, as we've counted, nine
>claims that do not reference "Bus"."

The smarmy attitude is helpful. Basically he is saying that
the > 100-headed hydra may be left with < 10 heads still attached?

>"Apart from the "Bus" issues, there are still a number of separate claims
>that Infineon has to prove they did not infringe. It will only take one
>claim in Rambus' favor to win the case."

Two words come to mind: pyrrhic and victory.

If Rambus comes out of this with only a few claims still standing, it
will probably be easy to work around them and more to the point it
will probably be cheaper to do so than to pay the extravagant license
fees that Rambus currently charges for SDRAM (to favor RDRAM of
course). E.g., if the programmable timing claims stand then Infineon
and others could go back to fixed timing.

This piece is like the investor sites that on Thursday decried the
Rambus slide saying that Electronic News and The Register couldn't
possibly know what the judge was deciding, quoting Rambus sources
sort-of denying the story, and mentioning the previous drop that
followed Barrett's comments only to recover quickly. They seem to
have omitted a few mea culpa's that were in order on Friday.

<FONT color="#0000FF">
\green{Happy St Patrick's day!}
</FONT>
Esc-x greenify-buffer

Rob.



This archive was generated by hypermail 2b29 : Fri Apr 27 2001 - 23:14:18 PDT