Decades-Old Rambus Litigation Against Micron For RDRAM Tech Reaches Settlement 82
An anonymous reader writes "The decades-old Rambus litigation against Micron for RDRAM tech finally reached a settlement. RDRAM tech has already been licensed by NVidia and Broadcom and has been used in game consoles such as the Nintendo 64. The preliminary deal is to last 7 years and net $280M for Rambus and Micro to gain access to patent licenses defining the technology."
Re:Now I feel old. (Score:5, Informative)
Statute of limitations only applies to when a case can be filed. A trial can go on for years past the statute.
Re:Now I feel old. (Score:4, Informative)
Exactly,
Join a consortium to develop a standard, sneak out at a quarter to four and patent the whole thing behind everyone's back and start throwing sue balls at everyone. They never had an original idea, or a working product.
This is horrible news (Score:5, Informative)
This is absolutely disgusting. Rambus is the ultimate patent troll. For those not familiar, here is some history.
Back in the 90's, Rambus became a member of JEDEC, an industry consortium of RAM manufacturers. JEDEC rules require members to disclose any patents that are relevant to the technology being discussed. Rambus didn't. It sat in on the meetings, listened, and modified its pending patent applications to cover DDR RAM. After the new RAM standard was adopted, Rambus surfaced its submarine patents and started suing everyone left and right.
Add to that the fact is that Rambus itself does not manufacture anything -- it's a technology licencing house that has a few engineers and an army of lawyers -- and you get a perfect example of a patent troll.
Re:Now I feel old. (Score:5, Informative)
The Rambus debacle actually pre-dates Slashdot.
The patent cat fight started in 1990, and slashdot started in 1997. It was old news by the time it was first mentioned on Slashdot.
Re:Now I feel old. (Score:4, Informative)
This particular debacle started in 1990. In 2000, Rambus filed lawsuits against every memory manufacturer in a much larger debacle. That's the debacle that dominated Slashdot for months and contributed to a huge influx of users.
Re:Now I feel old. (Score:2, Informative)
Well, there was probably one or two optimizations in RD-RAM that seemed meaningful over "regular" RAM at the time (probably had to do with timing and read/write operations, though- didn't RD-RAM read on both up-tick and down-tick of clock cycles, or something like that?), and Intel thought it'd really make the Pentium 4 a powerhouse system. Except Intel really overreached with the P4 (and Itanium was developed around then too, right?), and forgot the most important part about RAM, that being the "Random Access" bit.
RD-RAM could feed data onto the bus serially rather fast, once it found it, but really sucked at latency with random access reads/writes. But that fast, high-latency streaming seemed to work well for P4's rather deep instruction pipelines too on some benchmarks, when it was in the sweet spots for them together (P4 resolving a deep instruction pipeline while RD-RAM retrieved memory and started pumping data onto the CPU bus, which was then ready to start reading it).
Meanwhile, it became evident that multi-core CPUs provided a much more responsive system than the high clock rate P4 systems, even if those cores were running at less than half the clock rate of the P4, and it seemed like the P4 should have kicked ass over other systems.
But then those stupid P3 hackers in Israel made a couple of improvements in the P3 core (my dual processor 833 MHz P3 box was far more usable than a 2GHz P4 with gobs of RD-RAM at the time), which begat the Core, Core 2 and Core 2 Duo processors, and the P4 just quietly faded away as the P3 technology got to move on.
So, in IT, every once in awhile the right idea/implementation wins out over someone's over-hyped/over-sold pet project...
Re:This is horrible news (Score:5, Informative)
Because US patent system effectively legalizes extortion. Some manufacturers rolled over and paid "licencing", but others, like Infineon decided to fight. The jury in Infineon v. Rambus ruled in favor of Infineon, but then Rambus appealed to the federal circuit and got that decision overturned. PJ, of groklaw fame, put it thusly: "federal circuit has never seen a patent it didn't like".
This is just a short summary, you can find more info if you like. Rambus extorted money from pretty much everyone who has so much as touched RAM.
Re:Now I feel old. (Score:4, Informative)
Re:Now I feel old. (Score:4, Informative)
You are assuming that the final manufactured product actually matched the Rambus design, rather than being a different design that just happened to be patent encumbered.
I only have hearsay from people that worked on this stuff, but it all agrees on the fact that Rambus did not have a working design.
Re:Now I feel old. (Score:5, Informative)
You're badly misunderstanding people's gripe with this whole thing. Of course the patents were valid. The whole point of the consortium was to come up with a non-patented standard, so when Rambus went behind everyone's back and patented the whole shebang of course there was no prior art. The other memory makers had gone out of their way not to patent it.
And FYI, JEDEC won against Rambus in their lawsuit - Rambus was guilty of patenting stuff which was being discussed in secret at JEDEC. The problem was Rambus was only guilty of violating the terms of membership for JEDEC. The lawyers who drafted the membership agreement neglected to put in any penalties for patenting stuff behind everyone else's back, and as a result the only recourse JEDEC had was to kick Rambus out. There were no federal or criminal charges because they didn't break government law, they broke JEDEC's rules. The patents were still valid though, even though Rambus effectively stole them, because the there was no prior art and the USPTO decided it was patent-worthy.
Speaking of which, do you really think DDR is worthy of a patent? Counting a clock tick not just when the voltage goes up, but also when the voltage goes down? That should be obvious to a 5-year old.