duplicate-nickname writes
"Finally some good news on the patent font. A US district court tossed out most of Rambus' patent infingement claims against Infineon. The judge also states that even if Infineon has infringed on the remaining patents, they didn't do so willfully."
Re:Three patents left... and two questions (Score:1)
All your RAM are belong to us.
Re:RAMBUS - no longer the memory maker (Score:1)
Re:Not uncommon (Score:2)
unless the liscensing agreements state that it does. In this case all or almost all of the dram manufacturers put clauses in that stated that no payments would be made if rambus lost a patent case. If any of these companies have paid rambus and rambus has to reimburse them then this may bankrupt the company. We can only hope...
So to recap... (Score:3)
Doh!
It sounds like a business method NOT to patent!
The most interesting thing I read about the case.. (Score:1)
- A.P.
--
Forget Napster. Why not really break the law?
Re:Rambus vs. Microsoft (Score:1)
That's the trouble with doing stuff in an open fashion - there's always someone who comes along who's just out to screw everyone else.
_____
Sam: "That was needlessly cryptic."
Re:RAMBUS - no longer the memory maker (Score:1)
Actual JEDEC Regs (Score:5)
First of all, basic rules. The legal guide [jedec.org] (and other documents) stipulate that members will participate in good faith (Section C.1) and no restraint of trade activities of any kind will take place (Section A.1), and that JEDEC will not police or audit its members or their reports to the organization (Section E). It's a pretty open organization.
Now for the heavy lifting. The JEDEC Manual [jedec.org] has a lot of interesting stuff in it, especially in Section 7.3 which deals with use of Patented Products. First and foremost, JEDEC does not care if patented tech is used, but it does require that the holder agree to grant licenses freely or for a reasonable fee. And as some people have been asking, members must disclose 'any knowledge they may have of patents, or pending patents, that might be involved in the work they are undertaking.' (emphasis mine) Also notice the word 'might', I assume this means if you're not sure, say so anyways, good faith and disclosure and all that. So all members are aware of this clause, it is quite important because the front cover of all publications must include a blurb about patent compliance, all tech referencing patented technology must be noted, and footnotes must say 'compliance with this section of the document requires the use of patent No. xxxx' (or patent has been applied for...), this is described in Section 7.3.1
FYI and draw our own conclusions (IANAL).
Micron started a price war (Score:2)
About 4 weeks ago, I think certain types of memory prices started tumbling. I'd just had a system specced out, and the next morning ddr memory fell through the floor. Initially, crucial.com (micron's retail subsidiary) wasn't merely below the best prices for generic memory on pricewatch (for generic!), but *way* below. It's about $140 for 256mb of ddr memory at the moemnt, regular or ecc..
.
also down is sodimm memory (for laptops). We got a pair of 256mb's for $187 each for this laptop from crucial; they'd been noticably higher a few days earlier..
.
hawk, economist, who loves price wars (but already moderated this thread and must post anonymously)
(but it unmoderated anyway, sohere it is without anonymity!)
Re:The most surprising thing (Score:2)
UNFORTUNATELY, none of those things can happen. Does anybody have any idea whether the JDEC agreement has any teeth? And if not, what good is it, and what freaking moron agreed to that? It's pretty obvious by now that these standards bodies and joint efforts MUST have strong legal bindings, because the stakes are way too high to allow any one player to take advantage of the market like this. With Microsoft and Netscape running roughshod over W3 (and pretty much every other standards body that made the mistake of getting involved), you'd think that people would have had a clue before entering into partnerships like that.
That's how Patents work. [Re:whoa! ???] (Score:2)
Because that's how patents work. Copyright implies that your work must be derived or copied from mine to be infringing. Patents however just say it must be the same thing produced at a latter date.
Simply put an illiterate blind man on a desert island can violate the patent of somebody he has never herd of. Doesn't matter if he never read the patent or saw a device from the author of said patent.
--
Quidquid latine dictum sit, altum viditur.
Whatever is said in Latin sounds profound.
Re:Not uncommon (Score:5)
Notice, however, the two companies that didn't cave when Rambus' lawyers came knocking? The two not based in Asia: Micron (USA) and Infineon (Germany). Rambus hoped that markets worldwide would suffer from Asia's hard times, sapping Micron's and Infineon's strength. Unfortunately for Rambus, Europe and Wall Street recovered very quickly, and Rambus got the last thing they wanted: A real, live, honest-to-God court battle.
Unfortunately for anyone who submitted to Rambus, this case probably won't change their situation. A failed patent defense doesn't automatically undo any licensing agreements already signed. IANAL, so I ask you this: What changes because the judge invoked RICO? Does that open the door for the Dramurai to have those license agreements revoked, or bring criminal charges against Rambus execs and lawyers?
We're not scare-mongering/This is really happening - Radiohead
Re:RAMBUS - no longer the memory maker (Score:1)
They don't actually produce DRAM, they simply license their RDRAM technology to DRAM makers, and have recently started charging fees for SDRAM production as well, based on their questionable patents.
It is possible to have a brilliant idea, patent it, and have everybody use it a few years down the line, and have them pay royalties for the invention.
It is NOT possible to own dozens of patents in an area where you haven't produced anything yourself.
Say, Micron, being in the industry for a decade might own 5, 10, 20 or 100 patents in the area. They are in the field, and they have full-time experts trying to improve on what they already can. And they can test what they think up. Lots of brilliant ideas end up unusable because the drawbacks that you find if you actually try to use it weigh out the advantages.
Now when I first heard about rambus, it was a neat idea. That was a one-man effort. And it's possible that they got enough funding to put 20 or 100 experts in a room and tell them "now go and invent something". But without them actually designing and making the chips, it is doubtful that they would come up with anything patentable.
Maybe one or two, but certainly not a dozen or more.
Anyway, that's what I think.
Roger.
Re:whoa! what does this mean??? (Score:3)
Re:Not uncommon (Score:1)
Wow, that's almost half the price the local shop was quoting me. I tend to buy my parts locally because it's very nice to have a store to walk into.
Re:Not uncommon (Score:2)
Hey, I'm an AMD fanboy and I caught that fundamental stupidity. DDR memory costs a fortune, and I know it doesn't have to, and that irritates me no end. I just bought a nice AMD system, and I would've gotten a DDR system if the memory weren't a little more than twice as much as non-DDR RAM.
These stupid patent things cost everybody to feed a parasite.
Re:Oooh! (Score:3)
"And we can't watch FOX because they own those chemical weapons plants in Assyria!"
--Homer Simpson
--
Re:The expertise of the judge (Score:2)
Three patents left... and two questions (Score:3)
And on a related matter, what could Infineon risk if they end up guilty of accidentally infringing on these patents?
The loosers lost (Score:1)
Re:whoa! what does this mean??? (Score:1)
Willful Infringement Enhances Damages (Score:4)
Not much of a surprise . . . (Score:5)
Rambus had a bunch of patents. Threw 'em all at the defendants. Wouldn't agree to limit the number of patents or claims prior to trial. At trial, plaintiff only addressed a narrow number of claims in a few of the patents. [It would be advocacy suicide to try to prove up a zillion claims from a zillion patents].
Unsurprisingly, after plaintiff rested its case, the defendants moved to dismiss all the counts for all the patents on the grounds that Rambus didn't meet its burden of going forward. The judge agreed with respect to all the claims of some of the patents, and some of the claims of the remaining. But several patents, and many claims, remain.
A defendant needs to infringe but one claim of one patent to be liable for infringement and subject to an injunction.
So, after half a trial, we have narrowed the issue to the remaining claims. But defendant's position on liability remains just as precarious as before.
Re:Rambus vs. Microsoft (Score:2)
You are mistaken if you think that a semiconductor manufacturer needs to own a fab.
Most semiconductor companies are 'fabless-foundries'. That is they design and sell the circuits, but outsource the manufacturing. (as opposed to selling or licencing IP only)
Re:RAMBUS - no longer the memory maker (Score:2)
AFAIK, they never did. Rambus is an IP company, which means it has a few engineers (maybe), a handful of patents, and an army of lawyers. They don't actually produce any tangible product.
Not buying products based on Rambus' technologies makes plenty of sense from a price/performance standpoint. An Athlon of moderate to high speed combined with DDR SDRAM is the fastest x86 setup you can get for the vast majority of apps. (There are a few (very few) apps which can run faster on a P4, but they're definitely outside ±3.)
Re:RAMBUS - no longer the memory maker (Score:4)
Is it? The last Intel processor I bought was a 12-MHz 286 back in '90 or '91. I've bought several different AMD and Cyrix processors since then and haven't regretted any of them. I'll allow that everything I've bought from the 386SX-25 up through the K6-200 was purchased mainly because it was the cheaper alternative, but the K6-2-300 and K6-III-450 were bought in part because they weren't Intel parts (and they deliver more than adequate integer performance; floating-point performance is OK, but I'm not a gamer). The 1.0-GHz Athlon I'm running now is faster, across-the-board, than any P!!! you can get, and gives P4s clocked 30-50% faster a run for their money. That the Athlon is not from Intel is an added bonus.
Damages (Score:4)
So if you infringe on someone's patent, make sure you don't have any paperwork saying you knew about it before.
Re:The most surprising thing (Score:2)
Don't get me wrong.... I hate RAMBUS too, but... It was only a patent application at the time, and they had no guarantee that it would be granted. (It would have been nice to disclose the application, though.)
hopefully.. (Score:1)
Re:Really OT... (Score:1)
Re:Not uncommon (Score:1)
Re:RAMBUS - no longer the memory maker (Score:1)
Re:Really OT... (Score:1)
You point out a big problem with current 'punishment' system. 5 years or 20, you need to be rehabilitated. But the emphasis is on punishment, not actually making prisoners better people. If we did try for rehabilitation, i don't think we'd see as many repeat crimials.
Re:Really OT... (Score:1)
You are mixing apples and oranges.
Before the arguement was between murder and attempted murder. In both cases, the intent was to murder; the only difference there was whether the person actually succeeded.
Now you are talking about the difference between murder and manslaughter. One is a purposeful attemp to end someone's life (which, if muder has succeded), the other is the accidental killing of another. There is not 'attemped manslaugheter' because your intent in a manslaughter case was not to kill anyone, and therefore almost killing someone BY ACCIDENT is not a crime.
The only contradiction here is in your thinking.
Re:Not uncommon (Score:5)
Seriously, though, *IF* AMD can buy out the guts of Rambus, then it would be tremendously funny :-).
--
Re:RAMBUS - no longer the memory maker (Score:2)
Maybe if you happen to like Intel's previous reign as the undisputed price gopuger of cpu's or you support their current place as most overrated & underperforming bunch of sniveling babies it might be hard... Otherwise it's pretty simple...
Re:Its about time... (Score:3)
big difference.
nonetheless, a step in the right direction.
Oooh! (Score:5)
Maybe they'll be investing in a more ethical company now, like International Cruelty to Children and Small Animals. Or Allied Chemical and Nuclear Waste Disposal in Poor Neighborhoods and Third World Countries...
Answers (Score:3)
IIRC, the term of patents is 17 years from filing date. The article didn't state what the filing date was on the three the judge has left in play, but assuming 1990 (the date I remember from earlier reports on this as being the filing date - although the patents were apparently rewritten severely much later, the filing date was kept) which is probably correct, they would expire in 2007.
IANAL, but it's my understanding that since the judge has already ruled there was no willful infringement, the worst they could get would be an order to pay the licensing fee to rambus, both on future production and on already shipped product.
In a case of willful infringement, they would face the same figure multiplied by three.
"That old saw about the early bird just goes to show that the worm should have stayed in bed."
Re:The expertise of the judge (Score:5)
So even tho' the subsequent applications use only the term "bus", that term must be interpretted to mean "multiline bus", which is technology clearly not used in SDRAM or DDR, as there are clearly distinct lines for address, data and command signals. Any, IANAL, but I have been following this case closely.
Really OT... (Score:1)
Sorry, just a peeve of mine...
Re:Really OT... (Score:1)
Now, if you die, I might spend 20 years in jail - maybe even be rehabilitated by that time. If you don't, maybe I spend 5 years, maybe even less, in jail. When I get out, I might be even more pissed off and want to kill more people because of it. I'm still the same person that was ready/willing/able to kill.
Re:Really OT... (Score:1)
I say, if that's true, then why should the intent to murder someone, the attempt as opposed to the actual murder, carry a different sentence? One part of the law is setting a punishment based on intent, and one is setting a punishment based on outcome of the intent. A contradiction.
Re:Not uncommon (Score:3)
Have you checked prices on DDR latley? Doesn't look too bad to me. [crucial.com]
Re:whoa! what does this mean??? (Score:2)
Even killing someone has diferent classes and penalties, depending on wether the action was willful (Murder) or accidental (Man slaughter), and so it is the case with patents (Or appears to be).
RAMBUS - no longer the memory maker (Score:3)
I don't know if there's an active boycott of Rambus products, but I think there should be. The industry should also do something about Rambus, if possible.
Re:On the what? (Score:3)
Re:The most surprising thing (Score:2)
Re:Rambus vs. Microsoft (Score:4)
So how did they steal from RAMBUS ?
Because... (Score:1)
So the pragmatic way is to judge on both what was likely to have actually happened and what was likely to have been intended.
BTW, if you're going on about inconsistencies in the laws, there are plenty. If you want a lighter sentence don't kill people with guns, kill em with a car.
Cheerio,
Link.
Re:whoa! what does this mean??? (Score:1)
on thier patent claims for ten years before
the claims were actually made.
So either Infineon was unwillingly infringing
on a patent that could itself get thrown out
for showing prior art existed, or Infineon has
to give up the plans for the time machine they
used to get a decade jump on the competition.
Re:The most surprising thing (Score:2)
IIRC, JEDEC rules require the disclosure of any patents, granted or pending, that relate to proposed technologies. Rambus was required to disclose any pending patents that might relate to SDRAM.
Re:Rambus vs. Microsoft (Score:3)
I think that the general Slashdot population is glad to see Rambus go down because there is a sound argument that Rambus, both the company and the technology, is bad for the industry.
Lets first look at the technology itself. It turns out that RDRAM is a terrible technology for main memory. It only shows advantages on certain classes of problems, and those problems aren't the ones that most users have to solve on a daily basis. Also consider that RDRAM is much more expensive to produce than normal DRAM, and you really begin to see a total disaster for the consumer.
From a business standpoint, having a single company controlling DRAM technology can't be good. Look how bad it is with processors. We're still using an architecture that should have been thrown out years ago because of the monopoly that Intel has on the commodity processor market. Do we really want to see the same thing happen with memory?
Re:On the what? (Score:1)
the patent font
TrueType enters the legal fray... damn.
PDHoss
======================================
Sense from the bench! (Score:3)
Re:The expertise of the judge (Score:2)
However, it is important to note that in the present system, there are plenty of judges who specialize in certain areas of the law. The courts are partitioned such that different kinds of cases tend to fall under the jurisdiction of judges who are most experienced dealing with them. There are judges that deal in bankrupcy law, small claims, drug enforcement, devorce, capital crimes, and so on. There are judges who specialize in patent and trade disputes, and among them there are judges that specialize in technical issues such as this one. I can't say anything about the particular judge presiding over this case, but I do know they exist.
However, it is also important to note that although they do exist, we don't have enough of them. It is also important to note that the system that places cases before courts experienced enough to deal with them is more or less informal. It would be silly to file your multi-billion dollar patent dispute case so that it would be presided over by Judge Judy. Any compotent lawyer would file the case with the appropriate court so that it will have a good chance of being heard by a judge with the right background. It benifits no one to have a clueless judge.
--
Et tu iudex (Score:3)
Re:Oooh! (Score:3)
8 cents!
That's quite a PE...
At any rate, what's silly is that they're the typical "We have a patent! We're a monopoly! Life is good! We'll be endlessly successful!"
While the gov't allows patent monopolies to encourage innovation, they have a little thing called expiration dates... in 17 years, you'd have better come up with a new and better patent to compete with or you're SOL.
So while 17 years is a long time, remember how quickly the computer industry changes. With RAMBUS's backstabbing of JEDEC, and going sue-happy, no one in their right mind would even talk to them again. Not that there's much prospect of them producing anything worthwhile anyways, but if they did (do they spend their license fees on R & D, or just lots of Ls), companies will think good and hard about even talking to them.
Welcome to the real world RAMBUS. You can screw a company over once, but don't count on doing it again...
Now can we laugh at other memory makers? (Score:1)
Re:Answers (Score:4)
You do not recall correctly. Currently, the term of patents is 20 years from the filing date. At one time, the law was 17 years from the date of grant, and for some patents it is currently "17 years from grant or 20 years from filing, whichever is longer." But 17 years from the filing date never entered into it.
For more details see here [spi.org].
Rambus vs. Microsoft (Score:1)
We all know and love Microsoft's practices of stealing other companies technologies and incorporating it into the great MS empire. Yet for some reason the Slashdot community seems to rejoice in the fact that memory companies stole Rambus' designs.
Consider the following docket [rambusite.com]from the trial which indicates a fairly serious conspiracy to take Rambus' designs and modify them only enough to avoid paying patents.
Do I believe Rambus should spend more time inovating and less time litigating? Of course. But they didn't have the 20 billion dollars to startup a manufacturing plant so they licensed the technology to a company that did. Shouldn't they be able to claim the 00.75% royalty that they are asking for?
The most surprising thing (Score:4)
What it seemed like to me was that there was *some* wrongdoing on the part of both parties, and what it really boiled down to were the motives of RAMBU$ -- why didn't they point out their patents so long ago when open standards were being created, and why did they wait so long to enforce them after they allowed it to happen?
Of course, this case very well could be crystal clear, and the RAM_US legal team just happens to have their own Steve Jobs Reality Distortion Field®.
On the what? (Score:4)
Is there a church of patents or something?
Re:whoa! what does this mean??? (Score:5)
Since Rambus haven't been doing too well, they could really do with the cash.
Re:RAMBUS - no longer the memory maker (Score:1)
The only exception so far has been my Playstation2, which I just simply couldn't live without.
--
Re:Not uncommon (Score:1)
I'm rather pissed off at this 686B problem as it is right now, and personally I'm not too keen on VIA because of it, and the fact that AMD doesn't want to produce their own chipsets. I'd like to see AMD succeed as well... but I'd like to see a widely-implemented chipset that rivaled the BX in stability and ease of use.
Re:Not uncommon (Score:2)
I think that you are missing the point, here. The "contract" merely states that, if Intel sells enough Rambus chipsets, it can buy a lot of Rambus shares at a very low price. But, if patent suits begin to go against Rambus, those shares aren't worth the trouble. Ergo, Intel gives Rambus the finger, and goes DDR or what it will damn please.
Just my $.0.02
Re:whoa! what does this mean??? (Score:3)
One of the other goals is, of course to get a broadly reaching patent. At any point a judge could rule that the scope of a patent is too broad and infringement would be penalized to a lesser extent. For this reason, it is in the best interest of companies to be careful about how far they reach to get a vary broad patent on a whole class of technology.
Also, I can't relally blame Rambus for trying to take back some of the ground they lost over the past two decades as memory became a commodity. This is just good business.
--CTH
--
What does all this mean? (Score:2)
What do you guys feel? Is basing and developing further on other's research illegal and if so should it be considered so?
It's been a long time coming... (Score:1)
Re:RAMBUS - no longer the memory maker (Score:2)
Re:whoa! what does this mean??? (Score:3)
whoa! what does this mean??? (Score:5)
Don't get me wrong, I hate all this patent shit too, but if someone infringes on a patent, since when does it matter if it was "willfully" or not?
Re:Damages (Score:2)
Mods: How can I be redundant when I was 2nd post? (Score:1)
Re:Rambus vs. Microsoft (Score:2)
Exactly. As I recall, JEDEC even didn't have any objection to a member proposing to USE patented IP in the standard, so long as the license was open and reasonable. However, the requirement of disclosure was REQUIRED so that the other JEDEC members could negotiate such agreements WELL in advance of the patented IP being included in SDRAM.
The need for this is obvious, as to do otherwise would allow one member an unfair competitive advantage over the others, and the ability to DICTATE terms to the others after the standard had already been adopted.
This is exactly what Infineon and Micron allege that Rambus did (and they have the internal memos to prove it). Also, Rambus used an inside source at JEDEC called "secret squirrel" AFTER they had left to glean more info about the proposed SDRAM standard.... This info they used to amend their patent applications so as to patent IP actually not their own, so as to ensure SDRAM would infringe with their patents.
Infineon has already provided the Judge with enough evidence to get him to invoke the RICO act fraud-abuse exception to "attorney client priviledge" to allow them to discover more evidence of this.
This ruling, setting aside all but three infringement claims, and those having had their teeth pulled by his ruling that it wouldn't be ruled INTENTIONAL infringement.
The patents those three remaining claims are based on are unlikely to hold, given the evidence that has already been presented, and the additional evidence that will likely be found by Infineon as they go to trial.
Re:Really OT... (Score:1)
Re:RAMBUS - no longer the memory maker (Score:1)
Don't be silly...it's been over 10 years since I last bought an Intel product. You can get AMD or VIA CPU's; mainboard chipsets by VIA, SiS, or ALi; graphics chipsets from Nvidia, Matrox, ATI, ST Micro, Trident, S3, etc; NICs from 3COM, NetGear, LinkSys, etc; network switches from 3Com, Cisco, and several others; managed hosting from a slew of ASPs...ad nauseum. You can't name one Intel product that the average or exceptional tech user would need that doesn't have a non-Intel alternative. Heck...a couple years back my company was buying new Compaq Deskpro 2000 business PC's, and they came with Matrox display adapters, TI NICs and VIA chipsets. The only thing in them that actually was Intel was the CPU. So even in the big business world their grasp isn't as strong as they'd like you to believe.
Also, I believe that SDRAM chip makers must pay royalties to Rambus. So, you'd have to boycott all SDRAM.
You believe incorrectly. None of the SDRAM chip makers must pay royalties to Rambus except those that signed licensing deals with them for SDRAM rather than fight the Rambus claims in court. That's what this whole lawsuit is about to begin with. (duh!) Right now it appears that the only two who aren't paying royalties to Rambus are Infineon and Micron. So just buy your SDRAM or DDR from Crucial Technology and you're safe (they are the direct-sales division of Micron).
Re:Rambus vs. Microsoft (Score:1)
Bah! I took you up on that offer and read through that docket. First off, it's from a proceeding that took place almost a month and a half ago, so it certainly qualifies as old news. Secondly, it certainly does not in any way indicate a conspiracy of any sort to take Rambus's designs. They quite plainly state that they would design a "public domain" version of the memory that Rambus was trying to sell them. They wanted to take some of the elements of the Rambus design that were not patented or patentable (in the public domain or obvious) and use them in a new design that specifically excluded patented info. They wanted to design a similar memory interface that wouldn't be an infringement. It hardly qualifies as stealing and there certainly is nothing wrong with it.
What Rambus alleges in this document is way up the scale from being even remotely shady. After all, that's the purpose of patent searches. Before you make a new product, you look at everyone else's patented designs and try to make sure that your product does not infringe those patents.
Maybe RAMBUS will go RAMBUST! (Score:1)
Right now there is no reason to compare RAMBUS vs. SDRAM on performance, because obviously you'll get better performance from SDRAM, since you can put in more than twice as much for the same amount of $$$.
Re:Now can we laugh at other memory makers? (Score:2)
Actually no, the other companies deals fall if Rambus loses. This is not usually the case, once you sign a license you are bound by the license. In this case however I have seen reports that the license is terminated if the patents are invalidated.
Re:The most surprising thing (Score:2)
Oh yes. One of the major PC manufacturers played similar games with the PC bus in the EISA consortium. The patents were invalidated by the judge because of that. The legal term is 'abuse'.
Infineon have a strong defense if they can show that the RAMBUS patents were abused in that way.
The fact that the judge has struck out most of the claims during the trial is very bad for RAMBUS. An earlier poster pointed out that lawyer tactics are to throw out as many claims as they can in case one sticks. However the downside of that strategy is that after the judge has found 94% of your case to be unsubstantiated you have a big credibility problem.
Rambus have to win the case to justify their market cap. If they lose the incentive to license RDRAM technology will be pretty small and royalty rates will collapse from the hundreds of millions expected to the hundreds of thousands the idea is actually worth.
Incidentally, neither Netscape nor Microsoft ever pulled a stunt remotely approaching the RAMBUS one at W3C (or for that matter anywhere else thaty I know of). The closest Netscape came is with the SSL patent that they gave a royalty free license to within a few weeks of it being granted. Netscape did practice 'drive by submissions' where the proposal to the W3C started rolling off the fax only minutes before the new navigator version went for download. Microsoft has never pulled a stunt at W3C that I know of.
and it is only half over. (Score:5)
Re:The expertise of the judge (Score:1)
You state that the judge only needs expertise in the legal interpretation of the patent language. However, you then explain that this case revolves around the meaning of "multiline bus" and how SDRAM or DDR does not use it. This seems technical to me. How is the judge to know that SDRAM and DDR do not, in fact, employ a mutiline bus? It may be clear to you, but would it be clear to him? Does a judge understand what the defendent would say about address, data, and command signals? From my understanding, to prove patent infringement, the case must be made that the defendant used technology patented by Rambus. A judge must know whether Rambus is right in saying that the technology has been used, or the defendant is right in saying that it has not? Am I correct in this?
Should a judge be expected to know these things? Should a judge know all technical aspects for all industries? If so, I sure wouldn't want to be one.
--
The expertise of the judge (Score:3)
Do judges which hear patent cases consult with 3rd party (hopefully objective) experts concerning the related technologies? If not, then I suppose it is up to the companies to educate the judge in order to prove their cases?
Just a thought: Would it benefit society if judges were specialized. I know lawers specialize to a certain extent, but do judges?
Just wondering.
--
Re:whoa! what does this mean??? (Score:1)
The sad story about RAMBUS being implemented wrong (Score:5)
My first reaction was "NO, the bane of our existence is now going to fester in high end systems as well!"
He calmed me down and explained on several levels why RAMBUS was superior in this application. And by no means did he ever advocate the use of RAMBUS on a PC, he stated that it would be too costly and Intel had implemented RAMBUS incorrectly.
From what I got out of the conversation was this: RAMBUS was a high speed bus, but it is only 8 bits wide. Sounds ridiculous at first. Apparently Intel decided that 2 channels per CPU would be enough. Practically speaking, especially with high memory bandwidth application, they were somewhat correct. But the day to day performance was no better. DEC has decided to move ahead, but they are planning to use 5 channels per CPU. They are going to do many interesting things with these channels, one of the things he mentioned was the use of memory striping with parity. The probability for a multi-bit error being uncorrectable in this configuration is huge - tending to make more sense for high availability systems.
I also asked him to show me the money, where are there RAMBUS chips with the densities that are not laughably low. He indicated that the EV7 would be able to physically have 64TB of memory. That would, from what I have seen, need a memory stick the size of a 1 foot ruler per slot. The densities required for this are being made to order for DEC.
RAMBUS is a vile company, they don't seem to do much but glean royalties off of others for producing the technology. They produced something that was implemented poorly and is not especially useful for the average Joe. It would be a fitting end to see it end up on the Alpha, outside of broad market appeal. Low volume sales is just what the saps at RAMBUS need. They have done a few things to ensure royalties, they have planted RAMBUS in consoles, tried to get PC buyers to use it. They would want the price of a microwave to go up by $100 bucks too, if they could stuff RAMBUS in there somehow.
Why don't these people stop suing and whining, go to some area of the world to setup a memory plant, and make this stuff. Is it so hard to manufacture, so difficult to have good yields that they have to grab the coat tails of memory giants to pull the fabrication off?
I think NEC and Samsung should just be given the rights to RAMBUS, and have this pest of a company dissolved into the annuls of history. And Intel, shame on you for being duped into doing something so silly next time implement things correctly.
heh (Score:1)
Hopfuly....... (Score:2)
I am so glad that the system works when it realy counts.
Not uncommon (Score:4)
Re:RAMBUS - no longer the memory maker (Score:2)
Rambus is an Intellectual Property (IP) company.
They don't actually produce DRAM, they simply license their RDRAM technology to DRAM makers, and have recently started charging fees for SDRAM production as well, based on their questionable patents.
Rambus gets its just rewards (Score:1)