"Elpida Memory Inc. and Rambus Inc. today signed a patent license agreement for SDRAM and Double Data Rate (DDR) SDRAM memory.
Elpida Memory, Inc. is the name of the new joint DRAM memory venture formally announced by NEC and Hitachi on September 28th 2000. NEC and Hitachi established this joint venture company in December 1999 with the intention to bring their DRAM products together under a single brand and also to cooperate on the development and design of future DRAM products. According to IDC, Elpida will enter the DRAM market in January 2001 with 11.5% market share, and to grow to more than 20% market share by 2003.
Under the agreement Elpida can sell SDRAM and DDR SDRAM directly to their systems customers and ensure a seamless transition of DRAM sales from the NEC and Hitachi brands over to the Elpida brand. NEC and Hitachi, the parent companies, both retain their previously announced licenses with Rambus for both SDRAM and DDR SDRAM memory and Controllers that directly connect to them.
Under the licensing agreement, the royalty rates for DDR SDRAM are greater than the RDRAM compatible rates. The agreement also includes royalties for SDRAM as well as a license fee for the entire agreement."
This is a post I want to question:
"The patents were filed in 1990
The JEDEC meetings were in the mid-90's
Jedec kept no meeting minutes and no membership rosters, apparently.
We don't even know the dates of the meetings, or if Rambus was even a member of JEDEC.
While various changes may have been made to some of the Rambus patent applications
while they were being approved (they were filed YEARS before the JEDEC meetings), we
have no way of knowing what the relationship was of those changes to the JEDEC meetings.
It is only speculation that Rambus incorporated items being discussed at JEDEC meetings
into the patents. And, even if this did happen (just for the sake of argument), it may not
have been material to the actual substance of the patent.
All that said, the JEDEC meetings NEVER produced a working open-standard SDRAM memory
ANYWAY. It remained for INTEL to develop an SDRAM standard that actually worked, and
this was a few years later still.
Further, SDRAM aside, DDR remains an entirely separate issue from SDRAM. Rambus has patents,
separate from the SDRAM patents, covering DDR that the memory makers are also trying to
invalidate. However, DDR was a MUCH later development for the memory makers other than
Rambus, and DDR was NOT discussed at the JEDEC meetings, while Rambus' patents covering
DDR were also filed from 1990 to 1992. So, even if the Jedec argument holds and the SDRAM
patents are invalidated, that won't, in and of itself, invalidate the DDR patents.
Finally, the courts presume that the patents are valid and the burden of proof is
on Micron, et. al., to PROVE -- in the legal sense -- that they are invalid.
It's not a "slam dunk", but it's not a 50-50 shot either. My money is on Rambus."
Could it be possible that Rambus is right about these patents or no?
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)
Discussion Intel current and future Lakes & Rapids thread
Facebook
Twitter