Slashdot Mirror
Preliminary Ruling Limits Scope of Rambus Patents
Tackhead writes "According to this article in Electronic News, Rambus - our favorite litiga, uh, innovators in DRAM technology - has been smacked with a preliminary ruling that limits their patent claims to RAM technologies involving a multiplex bus. The article goes on to quote a source who says that since neither SDRAM nor DDR use this technique, this ruling could lead to the invalidation of RAMBUS' patent claims on SDRAM and DDR. Of course, this is just a preliminary ruling, and it's only one court battle (out of at least three), but it looks like the Good Guys (well, at the guys whose business is based on making chips instead of suing chipmakers) just might be winning."
Rambus is pretty much screwed because of this. Okay, they still have a few more trials, but my guess is that they're going to lose, and obviously a lot of their investors do to (at this moment, they're down 25%).
Intel has made their intentions clear. They will not be using RDRAM for future chips and will likely not work with Rambus in the future. Intel's not real happy about the lawsuits and they're definately not happy about the price gouging by Rambus. It hurts Intel's business to have to rely on RDRAM. The memory is so expensive, people don't want to use the motherboards.
Rambus has about the worst PR of any tech company because of the suits and because of the prices. Personally, I doubt they'll recover. They've really put themselves into a bad spot and I don't think they can dig out, but that's just MHO.
Their stock declined 26% today.
Hehe. Ouch.
sig fault
On the trivial part:
Why can't the patent reviewer call ??SU, with a which has a department that is well known for the technology that the patent covers. The reviewer ask to speak to a professor in that department. He ask the professor how (s)he would solve a problem covered by the patent. If the professor can describe within a ten-minute cold call the method covered by the patent, then it is trivial.
I think this would kill most of the high tech patents being given out today, because the patents only describe the obvious way to do something with a computer. (ie, it's patentable just because it is done by a computer seems to be acceptable by the USPTO)
Aah, change is good. -- Rafiki
Yeah, but it ain't easy. -- Simba
His link is invalid. I suppose its the dynamic servers over at Bloomberg. You will see an article there that says there was NO ruling yesterday. The ElectronicNews was WRONG.
Burn Hollywood Burn
The clear inference is that the order has been leaked, and the leak is reasonably accurate.
"The Chip Connection Specialists. We Never Stop Innovating." -- Rambus.com
"Microsoft strives to produce innovative products..." -- Microsoft.com
For people learning english, they might think the word "innovation" means "to stick a honkin' corporate phallus up the customer's ass".
"One might be tricked into believing a preliminary ruling limited the scope of Rambus patents, when in reality, it only means that a preliminary ruling limited the scope of Rambus patents."
You are saying the same thing twice and saying they are different. In the first part of that sentence, you say:
a preliminary ruling limited the scope of Rambus patents
And then you say, in the same sentence:
a preliminary ruling limited the scope of Rambus patents
So, if we make
X = "a preliminary ruling limited the scope of Rambus patents"
and
Y = "a preliminary ruling limited the scope of Rambus patents"
Then obviously X = Y, but you are saying X != Y ???
'Psuedo program to illustate my point
If (X == Y) {
C = X
}
So, we could say "One might be tricked into believing [%= C %], when in reality, it only means that [%= C %].
WTF??
Actually, the price of RAM in general has taken a major nosedive lately.
:-)
I can now get 128 MB DDR-SDRAM DIMM's for under US$80! =:-0 And it appears that the RDRAM makers have a clue, too: 128 MB PC-800 RIMM's now sell for under US$150. (Thud)
Maybe the competition from DDR-SDRAM has finally forced the price of RIMM's down for a change.
Raymond in Mountain View, CA
Yup. RDRAM is pretty clearly a Rambus innovation, and I don't begrudge them their royalties on RDRAM sales.
There's nothing intrinsically evil about saying "we invent things and license the tech to people who want to build them". It's only when Rambus said "Oh yeah, all your RAM are belong to us!" that I think they crossed the line from being innovators into being litigators.
If you invent something, you can license it. My beef with Rambus is that I don't believe they had anything to do with the "invention" of SDRAM or DDR, and that they therefore have no legitimate claim to royalties on those products.
If Rambus makes a fortune because Intel ships a CPU/chipset combo that turns the performance potential of RDRAM into real-world advantages, and their technology gains marketshare from DDR, more power to 'em.
But if the world goes with DDR instead of RDRAM, then Rambus (IMHO) should either come up with something better than DDR and patent it, or stick a fork in itself, 'cuz it's done.
In this case, I am wondering what RAMBUS is thinking.. Why are they doing this? Are they truly just a bunch of jerks? Why do they feel that they have rights to some of this technology? I mean, certainly someone there thinks they have a case. The more I read /., the more I think that all these companies are evil demons trying to take away my liberties and other people's hard work. Is this truly an accurate assessment? Surely someone can speak for RAMBUS to explain why they think that they are in the right on this matter.
The DDR patent still may hold water, as IIRC it was specific concerning the placement of a register in the memory. I think the JEDEC/RICO prospects are interesting, considering conflicting information coming out.
--
A feeling of having made the same mistake before: Deja Foobar
Problem is that it didn't limit the SCOPE of any patent. It merely said that, most likely (remember its a preliminary ruling), RAMBUS's patent doesn't even apply in the case of SDRAM. No "limiting of scope" and no invalidation of any part of the patent, just a Judge being safe and agreeing with an "expert" that RAMBUS has no patent on SDRAM.
Burn Hollywood Burn
it doesn't quite shut down the Rambus litigation machine. Besides the totally bogus stuff, they do have a patent on variable CAS latency in the DRAM. One could argue that this was obvious, since everyone and her sister was using controller-side variable CAS latency from 'way back when. Regardless, they have the patent and it does seem to bear directly on SDRAM and DDR.
DDR II, on the other hand, has had it designed out. Dang.
Lacking <sarcasm> tags,
--
A feeling of having made the same mistake before: Deja Foobar
Having 10 patents myself, I know a little about this business.
"There goes their patent" is actually a better phrasing of it. You can patent anything unique you invent - the trick is to patent something valuable - e.g. something the market actually wants.
RAMBUS went on this strategy of suing SDRAM makers because their own technology turned out to be inferior, offering massively higher price with no matching performance gain.
If RAMBUS is unable to enforce their patent against SDRAM, then the patent is - from a business standpoint - worthless. All it could be used for is preventing unauthorized copies of RAMBUS memory, which most customers don't want anyway.
--
A feeling of having made the same mistake before: Deja Foobar
Just a few week's ago - whether due to the greed that got them into this mess, or the pride of not being able to admit a mistake, or whatever.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
A valid point?
--
A feeling of having made the same mistake before: Deja Foobar
The reason why Rambus has to win this lawsuit is plainly for survival. If you look at Rambus' latest quarterly financial report at http://biz.yahoo.com/e/l/r/rmbs.html , you'll find that royalties constotuted 77.3% of thir revenues in the quarter ended Dec 31. Without these royalties, Rambus' would be pulling in just a little over 1/5th of their current revenues. That would make Rambus the stock hugely overvalued even at current prices. If you belive Rambus is going to lose the lawsuit, it's probably still not too late to short the stock :)
RDRAM - PC600, 16 bit
SDRAM - PC133, 64 bit
600 x 16 = 9600
133 x 64 = 8152
Now imagine DDR SDRAM, which is almost twice as fast since it reads and writes at the same time. RDRAM is just not worth it.
The real juice in the story wasn't the headline, it was the allowance of the possibility of fraud charges against prominent Rambusteers. If a fraud prosecution were to take place Rambus would completely collapse.
A well-crafted lie appears unquestionable - Dama Mahaleo
Most moderators would agree, probably nobody modded it past 3, but if a few people with mod pages load the page at once they'll see a funny post at 2, so two or three independantly decide to mod it up. When they all do, they find it's at 5.
Moderation is destined to be a little inaccurate when there's a delay involved like that.
And that's despite all the moderators who read at 2+, Highest First. All they ever see is the high-rated posts.
IMHO having moderator points should put you into 0+, nested.
I wouldn't force 'oldest first', but if anyone moded something 'redundant' while reading in 'newest first' I'd slap them with a tuna.
When I read the ./ headline and article, my first thought was, "there goes their patent". Then I read the article and reread the headline and it made more sense. Rambus patent hasn't been touched, the Judge has merely sided with an "expert witness" in saying that Rambus's patent doesn't even apply to SDRAM, since SDRAM doesn't use the technology in the patent.
Burn Hollywood Burn
If you're in the business where you're licensing something like that you're going to buy more than 512MB of RAM. 2 CPU P3 server vendors recommend 2-4GB, that's $4000 (CDN $) in RAM. In a 4 or 8-way server you're going to want a lot more.
The prices don't work for RDRAM for *most* things. If you're looking for a fast single CPU to do little except stream data, RDRAM is faster. But if you're worried about latency which is the real issue in any non-contrived server example, then RDRAM doesn't cut it anyways. The extra cost just makes it worse.
If RAMBUS actually offered performance benefits in those areas, it might justify the price.
But, if you really care about performance, you don't use x86 and you don't run NT. Sun, SGI, Compaq (Alpha), and IBM are in business for a reason...
... sanity entering the US court system? isn't that one of the signs of the end of the world?
"I don't need a compass to tell me which way the wind shines." - Mr. Furious, Mystery Men
Well, even if RAMBUS loses this case altogether, there is still the competition between DDR and RDRAM. Who will win this battle is completely unknown.
Do consumers want or need a P4 with RDRAM? Will AMD continue to take market share, thereby boosting DDR sales? Will RDRAM prices come down? Will DDR chipsets finally ship in volume? And which technology really works better?
This battle is far from settled even when excluding the court case. Intel IS going to ship DDR product with the P4, but at a later date, conceiveably after RDRAM is reaching mass acceptance. They are keeping the door open, though. RAMBUS isn't going away soon, if ever.
Burn baby Burn!
Probably not. Micron, for instance, never paid Rambus a cent - that's why they're part of another lawsuit ;-)
If you signed a contract with Rambus that said "we'll pay you a royalty on our SDRAM sales between $NOW and $LATER", well, uh, you signed the contract. Tough.
Of course, if you signed such a contract, and Rambus' patents turn out to be invalid, odds are you're gonna drive a harder bargain when the contract expires.
(I suppose that if a judge rules that Rambus knew its patents were invalid at the time it wrote the contract, it may invalidate the contract altogether - but that would require proof of fraudulent intent, which is a much harder thing to prove than anything that's been seen up to this point. Personally, I think there'd be reasonable doubt - Rambus' goons wouldn't have gone on the path of attempting to license DDR and SDRAM unless they had reason to believe they could get away with it. Even if the patents are invalidated, given what's been made public so far, I think Rambus can still credibly claim that they thought their DDR/SDRAM land grab would stand up in court.)
From the article...
If individuals are convicted of fraud, it could mean more than a financial slap on the wrist, sources said.
"This means that people could go to jail. That's what this means," the source said.
And they should be thrown in jail. RAMBUS obviously committed fraud by going into the JEDEC standards meeting knowing that that the standard reached would infringe on their patents, not notifying JEDEC that they were patent pending on technoligies that were in the standard and then modifying their patent applications to cover even more of the agreed upon standard. Hell, RAMBUS management deserves more than jailtime; they deserve to be drawn and quarterd. Maybe this (throwing RAMBUS employees and lawyers) into jail would discourage others from attempting to follow RAMBUS's litigeous lead.
Well if you'd been following the technology of RAMBUS and actually READING the reviews that you mentioned, you'd know that while RDRAM has wonderful bandwidth, it has horrible latency, heat, power, and timing problems. They've known about all of these problems since RDRAM came out and they're painfully obvious now. It may have oodles of more bandwidth than sdram BUT if the latency is bad or if you can't power it or design a decent chipset/memory controller for it, it's really worthless except possibly at the high end server market where people are willing to pay oodles of $$ for extreme bandwidth with massive cooling solutions.
"Christ what a design! I could eat a handful of iron filings and PUKE a better emergency pump than that!"
If I'm remembering correctly it actually went like this:
All JEDEC members were supposed to tell of any patents involving the area under discussion ahead of time which might interfere with any standards set by the group... This was not done by Rambus in that their pending patents which we are discussing now were not disclosed (even though they were only 'pending' they still should have listed them).
On the other hand I'm betting the management/legal people within Rambus didn't even know some of their ideas had been used for some time & it's pretty shaddy that they only bother to mention this as their fledgling product RDRAM is suffering a bad introduction to the market... That is the truly shady part of the deal in my thinking...
we are all invisible unless we choose otherwise
I can't vouch for whether or not they changed the byline while you were typing it, but the byline at "http://www.electronicnews.com/enews/news/NewsInde x.asp" for March 14 now says "Expected Rambus Ruling Could Limit Scope Of Patents"...
It sure as hell didn't say that this morning, when I read the article and submitted the link to Slashdot.
Forces of Evil
Forces of Good
These aren't complete lists but they're enough to help yout get by. Whenever you see a story, post "Yay! (Force of good) rocks! (Force of Evil) sucks!" and you'll do fine.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
Apologies to /. for posting twice in response to the same post. But if you visited the site early in the day, check your browser's cache. If you have a copy for the original article (this morning's version with the stronger by-line), you may wish to save it somewhere on your hard drive.
IMNSHO, Electronic News should do the right thing and print a retraction of the original article.
Another thing that happens--or at least it has happened to me--is that a moderator will be reading along, and come across a post he/she thinks is worthy of modding up. Maybe it's only at two and should be a three or so. So she/he mods it up--but while that's happening, another moderator comes through and sees the same thing. Since neither of them have refreshed while they were reading, even if one of them mods the post up five minutes before the other, they both think they're just bumping it up to a 3--but since they both did it, it's now a four. Inadvertently over-rated.
No relation to Happy Monkey
1. Go to standards commities, listen to all the sugestions, and patent them in the hope that one day the suggestions are use. (which is what RAMBUS is alleged to have done by some)
2. Go to a standards commities and suggest your recently filed patent as a standard, in the hope someone will listen and add it to the standard. (which is what MS did with CSS)
In order for standards to be accepted or even used for that matter, there NEEDS to be full disclosure and trust at the comitties. Meaning: No Patents Allowed. Really, the EU, The US, and other interested parties should sign a treaty that agrees on a uniform language to prevent patents from going into or out of a conference. That and the definition of "non-trival" when applied to a patent should mean more than "something a high school grad wouldn't understand". Sure, I don't understand RAM design, but there are those who KNOW what non-trivial is and it seems RAMBUS has patented a non-trivial technology.
Burn Hollywood Burn
Rambus makes 77% of its revenue by lawsuits and extorting money from companies as "license fees" for technologies that they did not invent. "Evil" seems like a pretty characterization of Rambus to me.
Heh, heh -- I hope those royalty agreements are binding, so the wimps have to pay for their wimpiness even if the Rambus patent is invalidated. It'll teach'em not to be so wussy...
Damm -- I wonder how much Rambus has collected already from their patent skullduggery?
--
Scott Robert Ladd
Master of Complexity
Destroyer of Order and Chaos
All about me
And of course, The Register's take on the SDRAM/DDR SDRAM.
Assuming this whole thing blows up in Rambus' face, this would end the SDRAM subsidy of RDRAM, which you can expect to see suffer an awful fate. Makes you wonder what Intel is thinking at this moment, with it's finger stuck in the door jamb.
--
A feeling of having made the same mistake before: Deja Foobar
FWIW, I leeched the headline pretty much verbatim from the Electronic News article. (IANAL, and the Electronic News article may have overstated the case, and I may have propagated the overstatement)
Your phrasing - "most likely" their DDR/SDRAM claims will not apply - is probably the more fair take on the story.
http://quote.bloomberg.com/fgcgi.cgi?T=marketsquot e99_news.ht&s=AOrD1LBRvUmFtYnVz
It's already been confirmed (even by infineon's lawyers) that there's been ABSOLUTELY no ruling and that the EBN article is false.
This is just more emulex-ish crappola to drive Rambus down.
Hasn't anyone noticed that Micron is one of EBN's partners? They'll do anything to knock rambus down. There are no "good guys" in this business. You idiots who consider the anti-rambus companies the "good guys" are seriously misguided. They're all in this for your money. Stop making holy wars out of monetary battles.
Hyundai is presently in rather dire financial straits due to a high debt load. Hyundai may just have been hedging their bets, as they may not be in a position to afford the cost of losing the case against Rambus.
It's also possible that since RDRAM is more expensive than SDRAM or DDR, and we all know how low RAM prices have gotten lately, Hyundai may have decided there's no money to be made in anything other than RDRAM. (And that there's therefore no money to be saved by trying to cut Rambus out of their SDRAM and DDR sales, since they may not be making money on these types of RAM anyways)
What does the former East Germany have to do with RAM?
Nothing to my knowledge, but I can see how you'd think that (Deutsch Democratic(sic) Republic). This is what all that alphabet soup really means:
I hope this clarifies what "DDR" means. Of course, I could be talking out of my proverbial ass.
All your hallucinogen are belong to us.
Will I retire or break 10K?