Rambus coming back?

[Ice9]

Interesting that Anandtech had the news of Rambus' newest lawsuits claiming copyright infringement on DDR2, GDDR2 & GDDR3... But they didn't have the even more significant news that Rambus was granted summary judgement against Hynix in their most recent court case:

http://www.forbes.com/technolo...artner=yahoo&referrer=

Now, BEFORE you reply to this, know the facts about Rambus. I'm a long term investor (and believer), and I know the ENTIRE story of what happened to Rambus, and how the memory manufacturers are trying to screw this tiny company. I also ultimately know that Rambus can't possibly lose in court long-term, since they invented the very technologies in use by most memory to date - and they did it back in 1988, four years BEFORE they were brought into JEDEC.

Also, if you think Rambus is "dead", you might want to know that they're supplying the RASER cell going into most (if not all) PCI Express implementations. They are also the predominant technology in the upcoming "Cell" technology, which leverages Rambus's Redwood and XDR technologies - as well as Flexphase.

 

[VirtualLarry]

Originally posted by: Ice9
Interesting that Anandtech had the news of Rambus' newest lawsuits claiming copyright infringement on DDR2, GDDR2 & GDDR3... But they didn't have the even more significant news that Rambus was granted summary judgement against Hynix in their most recent court case:

http://www.forbes.com/technolo...artner=yahoo&referrer=

Now, BEFORE you reply to this, know the facts about Rambus. I'm a long term investor (and believer), and I know the ENTIRE story of what happened to Rambus, and how the memory manufacturers are trying to screw this tiny company. I also ultimately know that Rambus can't possibly lose in court long-term, since they invented the very technologies in use by most memory to date - and they did it back in 1988, four years BEFORE they were brought into JEDEC.

Also, if you think Rambus is "dead", you might want to know that they're supplying the RASER cell going into most (if not all) PCI Express implementations. They are also the predominant technology in the upcoming "Cell" technology, which leverages Rambus's Redwood and XDR technologies - as well as Flexphase.

(emphasis mine)
LOL. Isn't that a little one-sided? Weren't most of the original lawsuits, filed by Rambus, against the other mfgs, exactly the opposite (eg. R wanted to screw everyone else). I don't blame the rest of the mfgs for trying to get rid of them. IMHO, RAMBUS should just dry up and blow away, the market doesn't want nor need them. They've set the industry and technology backwards by quite a bit by their antics. We would probably be using open-JEDEC-standard DDR4 on our desktop machines right now, had not Rambus colluded with Intel to strong-arm the entire industry in an attempt to force it to use their proprietary memory types, as deficient as they were.

I admit, their XDR stuff is probably useful, but could have just as easily come from another non-evil company, eventually.

 

[Terumo]

Now when will that transcend into a faster RDRAM module? lololol

My 32bit 1066mhz RDRAM is getting dated and sure would like a faster solution -- well, maybe not as the P4T533 is a power hog to begin with!

 

[Ice9]

Well, let's see. Read this, and poke holes in it as you see fit - but please view it objectively.

First of all... Rambus was formed by 2 people, Drs. Farmwald & Horowitz. It was their invention that led to the forming of Rambus. Their invention was back in 1988, and it was this invention that's the very base principle outlining the operation of SDRAM, DDR & DDR2/GDDR, etc

So, Rambus decided they did NOT want to be in the manufacturing business. Rather they'd settle on a royalty model, and collect a 2-2.5% royalty on every unit sold. In the interests of striking a deal, they showed their technology under NDA (non-disclosure agreement) to several companies while they filed their patents on their inventions. Samsung, Infineon (then Siemens), Hyunday (now Hynix), Micron, etc all knew about Rambus technology.

Then Intel started jabbering about how they were looking into Rambus technology at a JEDEC conference because it was faster than anything else out there, and they couldn't wait for JEDEC to come up with something. So, the involved companies at JEDEC (the same companies who were under NDA to Rambus) invited Rambus to participate...

Then they started the Cherry-picking for SDRAM. Intel did not COLLUDE with Rambus to strong-arm an industry - in fact, the very statement kinda looks silly when you look at the fact that Rambus is DEPENDENT on ram manufacturers. All they do is collect a royalty on their technology.

Anyway, the cherry-picking kept on going - so what did Rambus do? They amended their patent CLAIMS (not the technical portion of the patent) to cover what JEDEC was trying to do (outright intellectual property theft).

Now, to consider Rambus evil, you have to consider the ENTIRE patent process evil. Rambus created the technology that SDRAM and DDR use today - almost ALL of it. However it's the PATENT CLAIMS that were "in dispute".

Look at it this way. You're the guy who invented "TUMS". Your patent is the formula to make it. Your patent CLAIMS are what you use it for, for instance "For the temporary relief of heartburn". In the patent system, you CANNOT change the patent. That stays the same way forever and ever amen.

However, the US Patent system allows innovators to update their claims when they discover their invention goes beyond the original set of claims. It's not only LEGAL, it's ENCOURAGED.

In this case, let's say you discover that TUMS is being used as a calcium supplement to prevent osteoporosis. You are then ENCOURAGED to update your PATENT CLAIMS to include an additional claim of "For supplemental calcium in the prevention of Osteoporosis".

Rambus knew JEDEC was picking their patents, but at the time their patents were still pending. The law states you're not allowed to discuss your patents openly if they have not been issued yet. Rambus stayed silent the entire time because legally at the time they were not allowed to discuss their pending patents. However, the law allowed them to update their patent claims - which they did and were ENCOURAGED to do.

So tell me where Rambus was "evil" or "wrong" here again?

 

[Ice9]

Originally posted by: Terumo
Now when will that transcend into a faster RDRAM module? lololol

My 32bit 1066mhz RDRAM is getting dated and sure would like a faster solution -- well, maybe not as the P4T533 is a power hog to begin with!

RDRAM is dead, even Rambus says so. They've moved on to XDR and chip-to-chip interconnects, like Redwood.

And yeah, sure virtuallarry, maybe someone else COULD have come out with technology just as good as XDR, Redwood and Flexphase - but the point is no one did. Rambus did, they did it before anyone else, and they deserve all the royalties that result.

Just because they were first doesn't make them "evil". Intel, AMD and everyone else's favorite tech companies collect royalties on their inventions all the time - what makes Rambus attempting to do it any more "evil" than any other company trying to do it?

JEDEC wants you to believe that DDR & DDR2 belong to them. It doesn't. Get used to it.

 

[Zebo]

Real good info ICE, thanks... I bet your stock will do fine. I heard lots of stuff recently about video ram to be rambus..who knows all may go full circle back to rambus

 

[DerKaiser]

Rambus created the technology that SDRAM and DDR use today - almost ALL of it.

That is highly debatable and exactly what is at the heart of all the lawsuits.

 

[Ice9]

Originally posted by: Zebo
Real good info ICE, thanks... I bet your stock will do fine. I heard lots of stuff recently about video ram to be rambus..who knows all may go full circle back to rambus

With any luck, that'll be true. I never understood why people stand up for JEDEC and the whole "open standards" farce. JEDEC is a sloth compared to Rambus. People somehow associate "Rambus" with "Expensive", when it's Rambus that doesn't even set the pricing, only the royalty rate.

People believed that Rambus RDRAM was hugely expensive to manufacture, and in return, costly to consumers - then BOOM! Michael Dell calls a foul to all the DRAM companies claiming they're colluding to keep prices high.

hmmmmm so let's see... these JEDEC monkeys are colluding in an "open standards" forum to cherry-pick inventions from a tiny company that had a great idea. These same monkeys are ALSO colluding in the private sector to keep ram prices high overall at a time when it's most profitable to memory manufacturers... Then these same manufacturers hire smear campaign companies like Inquest's Bert McComas to make RDRAM look bad to everyone.

Oh yeah, the only LOGICAL thing to do when knowing the above is to hate Rambus! LOL It's all their fault!

 

[Ice9]

Originally posted by: DerKaiser

Rambus created the technology that SDRAM and DDR use today - almost ALL of it.

That is highly debatable and exactly what is at the heart of all the lawsuits.

No, it's not debated at all.... The heart of these lawsuits is willful patent infringement because of the patent CLAIMS involved, not the patent itself. These companies are trying to say that their invention doesn't infringe on Rambus' patent - but obviously that doesn't fly since Hynix was just found to be infringing in summary judgement. So much for that.

Again, this isn't about the PATENT. It's about the PATENT CLAIMS. Rambus owns the patent on the technologies used in DDR and DDR2. It was JEDEC's CLAIM that rambus used the open standards forum to amend their patent claims to cover SDRAM and DDR - which is an encouraged practice according to patent law.

Obviously JEDEC's defense won't work, because it's already documented that these companies were under NDA before Rambus was in JEDEC. The companies involved have one tactic and one tactic only - delay them with countersuits, stays and the like for as long as they possibly can in an attempt to bankrupt Rambus so they don't have to pay. It's not a matter of IF, it's a matter of WHEN.

On the plus side, Rambus continues to be a profitable company, and they haven't stopped innovating. Their stuff is in PCIE, HDTV's, and the upcoming Cell which will be used in the PS3 among other things. I'm not worried about Rambus long-term.

 

[VirtualLarry]

Originally posted by: Ice9
Well, let's see. Read this, and poke holes in it as you see fit - but please view it objectively.
First of all... Rambus was formed by 2 people, Drs. Farmwald & Horowitz. It was their invention that led to the forming of Rambus. Their invention was back in 1988, and it was this invention that's the very base principle outlining the operation of SDRAM, DDR & DDR2/GDDR, etc

Which invention?

Originally posted by: Ice9
Then they started the Cherry-picking for SDRAM. Intel did not COLLUDE with Rambus to strong-arm an industry - in fact, the very statement kinda looks silly when you look at the fact that Rambus is DEPENDENT on ram manufacturers. All they do is collect a royalty on their technology.

I guess you missed the entire Willamette P4/Socket 423/i850 era??? And yes, Intel + Rambus colluded - how can you conclude otherwise looking at some of the details of their arrangement? Btw, let me give you a hint, RAM mfg's are dependent on a market segment to sell to, and the largest is desktop PCs, of which Intel is largely the primary leader in controlling the direction that the technology takes.

Originally posted by: Ice9
Anyway, the cherry-picking kept on going - so what did Rambus do? They amended their patent CLAIMS (not the technical portion of the patent) to cover what JEDEC was trying to do (outright intellectual property theft).

LOL. That's certainly an interesting twist to everything that I've read about the story. Yes, I'm sure that Rambus is the victim here, attempting to hold the entire industry hostage, due to patents filed with amended claims based on the ongoing "open" discussions of a standards-body. Kind of like someone partitipating in the W3C attempting to patent hyperlinking on the side...

Originally posted by: Ice9
Now, to consider Rambus evil, you have to consider the ENTIRE patent process evil. Rambus created the technology that SDRAM and DDR use today - almost ALL of it.

That's ... an interesting claim. I don't buy it.

I noticed that you didn't mention anything specific here, either above, when you mention the "invention", or here. What exactly did they "invent", that is so core to current DRAM technology?

Originally posted by: Ice9
However it's the PATENT CLAIMS that were "in dispute".
Look at it this way. You're the guy who invented "TUMS". Your patent is the formula to make it. Your patent CLAIMS are what you use it for, for instance "For the temporary relief of heartburn". In the patent system, you CANNOT change the patent. That stays the same way forever and ever amen.

Do you understand what a patent is? You can amend the patent. The entirely industry hates "submarine patents", mostly, unless one happens to be an "IP-only" company, that makes their living off of suing everyone else, like Rambus, SCO, and a few other "evil" companies.

Originally posted by: Ice9
However, the US Patent system allows innovators to update their claims when they discover their invention goes beyond the original set of claims. It's not only LEGAL, it's ENCOURAGED.

I wouldn't go so far as to say "encourage". More like it's allowed, legally, but that's about all.

Originally posted by: Ice9
Rambus knew JEDEC was picking their patents, but at the time their patents were still pending. The law states you're not allowed to discuss your patents openly if they have not been issued yet. Rambus stayed silent the entire time because legally at the time they were not allowed to discuss their pending patents. However, the law allowed them to update their patent claims - which they did and were ENCOURAGED to do. So tell me where Rambus was "evil" or "wrong" here again?

That's completely BS. Rambus amended their existing patent filings, based on information that they obtained from the JEDEC forum, and never informed them (at the time) that they were filing additional patent claims based on that info, at least according to what I've read on the subject.

There's no law that says pending patent applications cannot be disclosed, in fact, it's often encouraged to announce. What's not encouraged, mainly to protect the patent-pending technology, is openly discussing the "secret details" contained within, basically telling how it works inside. But announcing that a patent has been filed, especially when you are participating in a standards-body forum, is the proper thing to do. Rambus withheld that information, instead (apparently) planning to hit up those very same companies some years later with lawsuits, for profit. (It is my opinion that they did this on purpose, because they knew that if the companies forming the standards, found out that the tech might have been patent-encumbered, then they might have changed the standard to work around that, and then Rambus wouldn't have been sitting-pretty a few years down the line, all set up to effectively legally extort an entire industry.)

It may have been legally allowed by the patent office, but in every other respect, it's sleazy and downright un-ethical. The industry doesn't need companies like Rambus. "Lawsuits for profit" should be an illegal industry, like selling slaves or babies. A productive and civilized society has no need of it whatsoever, it benefits society in no way.

 

[VirtualLarry]

Originally posted by: Ice9
With any luck, that'll be true. I never understood why people stand up for JEDEC and the whole "open standards" farce.

Yeah, let's all have a huge assortment of proprietary, incompatible, standards instead, and be forced to pay arbitrary, highway-robbery prices for them. Let's be forced back into a tech "dark ages", where no-one can interoperate with anyone else, and our computer still have 16KB of RAM, because another 16KB costs $10,000. Let's destroy the internet, who needs "standards, no-body benefits from standards.. I mean, can you imagine how much money you could make, to patent the use of the internet.

I thank the Lord above that Rambus wasn't involved with the W3C, and that I'm not paying a per-byte royalty charge to post this, an "IP tax" on IP protocol for instance.

I mean, what the heck does it matter, if the rest of the industry is in the slump of the Dark Ages, as long as Rambus the tech. feudal king is sitting pretty, taxing everyone else. You sup at the king's table, so you don't care that the peons are enslaved. "Let them eat cake." So to speak. Don't you feel the least bit of guilt for being a stockholder, knowing how un-ethically Rambus has behaved as a company?

Originally posted by: Ice9
JEDEC is a sloth compared to Rambus. People somehow associate "Rambus" with "Expensive", when it's Rambus that doesn't even set the pricing, only the royalty rate.

Are you ignorant? In a commodity market, attaching an additional royalty payment rate, surely increases prices, and effectively sets a minimum base price for the item.

Originally posted by: Ice9
People believed that Rambus RDRAM was hugely expensive to manufacture, and in return, costly to consumers - then BOOM! Michael Dell calls a foul to all the DRAM companies claiming they're colluding to keep prices high.

It wasn't expensive to mfg, at least not much more than other types of DRAM, AFAIK. It was the royalty/ransom payments to Rambus that made it expensive. Which is what would have made it hard to compete, price-wise, in the market against other royalty-free commodity memories. So that's why Rambus colluded with Intel to strong-arm the market, to effectively declare the need for the new Rambus memory by fiat, while at the same time, launching lawsuits against all of the other non-Rambus commodity memory mfgs.

Originally posted by: Ice9
hmmmmm so let's see... these JEDEC monkeys are colluding in an "open standards" forum to cherry-pick inventions from a tiny company that had a great idea. These same monkeys are ALSO colluding in the private sector to keep ram prices high overall at a time when it's most profitable to memory manufacturers... Then these same manufacturers hire smear campaign companies like Inquest's Bert McComas to make RDRAM look bad to everyone.

I'm not going to claim that there wasn't some collusion in the industry to keep DRAM prices high, I think that many suspected as much, but what Rambus attempted to do to the industry is: 1) unrelated, and 2) much, much worse. And they're not "monkeys", they're engineers, and they are probably the ones least likely to engage in price-fixing collusion, that would be the marketing arms of those various DRAM mfgs if anyone.

Originally posted by: Ice9
Oh yeah, the only LOGICAL thing to do when knowing the above is to hate Rambus! LOL It's all their fault!

No, I would never have thought that Rambus was responsible for any price-fixing going on between the non-Rambus DRAM mfgs. But it's interesting that you drag out that straw-man into the discussion, to attempt to make Rambus look good, by making non-Rambus mfgs look bad. Nice try. That doesn't make Rambus any less evil and unethical themselves.

 

[Ice9]

Originally posted by: VirtualLarry
Which invention?

There are several. The original '703 filing is the one that's most prevalent in each of these court cases. You can search for "Rambus" on the USPTO's patent reference. Look for the patent that ends in 703. This is only one of the 14 or so patents in which SDRAM, DDR & above infringe.

Originally posted by: VirtualLarry
LOL. That's certainly an interesting twist to everything that I've read about the story. Yes, I'm sure that Rambus is the victim here, attempting to hold the entire industry hostage, due to patents filed with amended claims based on the ongoing "open" discussions of a standards-body. Kind of like someone partitipating in the W3C attempting to patent hyperlinking on the side...

"Twist"? I read court transcripts and dockets. What are you reading?

Rambus IS the victim, and the evidence proves it. And how is Rambus trying to hold the industry hostage? Explain specific ways they're trying to do this? They're simply trying to get the royalties they are due. Do you think Intel would allow their property to be used without being compensated for it?

Originally posted by: VirtualLarry

I noticed that you didn't mention anything specific here, either above, when you mention the "invention", or here. What exactly did they "invent", that is so core to current DRAM technology?

Read the '703. Read the court transcripts. Read the FTC docket. Read the Infineon dockets. Read the Markman ruling and learn the laws of Kingsdown instruction. Educate yourself, and see what both sides said. You won't do so though, because you, like many others, are far more happy hating Rambus when they shouldn't be the target of your dismissal.

Basically, stop reading smear sites, and start reading actual law

Originally posted by: VirtualLarry
Do you understand what a patent is? You can amend the patent. The entirely industry hates "submarine patents", mostly, unless one happens to be an "IP-only" company, that makes their living off of suing everyone else, like Rambus, SCO, and a few other "evil" companies.

No. The technical abstract of a patent cannot be changed. If you change your invention, it goes into a completely new filing. Only the claims can be amended. Regardless, the patent did not change from the original filing (though it was sent back and asked to be broken up into several filings since the original filing contained "too many inventions" for a single filing).

Anyway, your IP-only scenario is not what Rambus is. They are a technology company that develops intellectual property. They have no desire to manufacture or warehouse anything. They are a small company that has 200 employees, 170 of which are engineers, the rest clerical & managerial staff. What Rambus has is a bona-fide invention that came way before JEDEC's SDRAM. And RDRAM, for the duration of its lifespan, proved to be a superior technology to SDRAM and DDR. It wasn't until RDRAM was forced out of the market that DDR finally improved - albeit marginally.

Originally posted by: VirtualLarry
I wouldn't go so far as to say "encourage". More like it's allowed, legally, but that's about all.

No. It is encouraged. It is an accepted practice. It happens every day, it's a standard component of patent law. What is DISCOURAGED is to make these claims overly broad, which is not the case with Rambus. In an patent dispute, the judge has the right to restrict the scope of any given patent claim that has prior precedent.

Do YOU understand what a patent is?

Originally posted by: VirtualLarry
That's completely BS. Rambus amended their existing patent filings, based on information that they obtained from the JEDEC forum, and never informed them (at the time) that they were filing additional patent claims based on that info, at least according to what I've read on the subject.

That is the case today. That was NOT the case when Rambus attended JEDEC.

Now... Tell me about JEDEC's patent disclosure policy? The patent disclosure policy did NOT apply to patent applications at JEDEC, nor could it have since it was not legal to do so at the time.

Now... Tell me, if you were Infineon, Hynix or Micron, and you were under NDA with Rambus to manufacture their technology, wouldn't it be reasonable to assume you knew about it?

So, if RAMBUS was not permitted by law to disclose patent applications, and they were under no disclosure policy at JEDEC to disclose, and IBM, HP and Intel ALL declined to disclose THEIR patents (which was OK with JEDEC), where was Rambus in the wrong again?

Originally posted by: VirtualLarry
There's no law that says pending patent applications cannot be disclosed, in fact, it's often encouraged to announce. What's not encouraged, mainly to protect the patent-pending technology, is openly discussing the "secret details" contained within, basically telling how it works inside. But announcing that a patent has been filed, especially when you are participating in a standards-body forum, is the proper thing to do. Rambus withheld that information, instead (apparently) planning to hit up those very same companies some years later with lawsuits, for profit. (It is my opinion that they did this on purpose, because they knew that if the companies forming the standards, found out that the tech might have been patent-encumbered, then they might have changed the standard to work around that, and then Rambus wouldn't have been sitting-pretty a few years down the line, all set up to effectively legally extort an entire industry.)

So, you're saying that even though these companies KNEW about Rambus' technology, and KNOWINGLY voted it into the SDRAM standard having SEEN this technology under NDA, it was RAMBUS's fault for not saying anything about it when they weren't even legally obliged to do so?

That's pretty funny!

It may have been legally allowed by the patent office, but in every other respect, it's sleazy and downright un-ethical. The industry doesn't need companies like Rambus. "Lawsuits for profit" should be an illegal industry, like selling slaves or babies. A productive and civilized society has no need of it whatsoever, it benefits society in no way.

 

[Megatomic]

deja vu

I've seen this thread happen before...

 

[DerKaiser]

No, it's not debated at all.... The heart of these lawsuits is willful patent infringement because of the patent CLAIMS involved, not the patent itself. These companies are trying to say that their invention doesn't infringe on Rambus' patent - but obviously that doesn't fly since Hynix was just found to be infringing in summary judgement. So much for that.

Patents are declared invalid regularly. Reasons for invalidating patents include the existence of prior art, the purposeful exclusion of inventors, etc.

You obviously don't understand the ruling that was made in the Rambus vs Hynix case. Hynix wanted a summary judgement saying that they weren't infringing any of Rambus's patents and dismissal of the suit. That motion was denied and now the arguments begin as to whether Hynix infringed or not. Hynix was not "found to be infringing in summary judgement". Basically, the trial to determine whether Hynix infringed can now begin. Shouldn't you take this crap to the Rambus message board on Yahoo?

 

[Ice9]

Originally posted by: VirtualLarry

Yeah, let's all have a huge assortment of proprietary, incompatible, standards instead, and be forced to pay arbitrary, highway-robbery prices for them. Let's be forced back into a tech "dark ages", where no-one can interoperate with anyone else, and our computer still have 16KB of RAM, because another 16KB costs $10,000. Let's destroy the internet, who needs "standards, no-body benefits from standards.. I mean, can you imagine how much money you could make, to patent the use of the internet.

Right. So, because rambus only invents the underlying technology and LEAVES IT TO THE MANUFACTURER to develop the physical interface, that contributes to its incompatibility? That's a stretch. A HUGE stretch.

I thank the Lord above that Rambus wasn't involved with the W3C, and that I'm not paying a per-byte royalty charge to post this, an "IP tax" on IP protocol for instance.

If it meant your internet connection would be 3x faster for a marginally higher cost, why not?

I mean, what the heck does it matter, if the rest of the industry is in the slump of the Dark Ages, as long as Rambus the tech. feudal king is sitting pretty, taxing everyone else. You sup at the king's table, so you don't care that the peons are enslaved. "Let them eat cake." So to speak. Don't you feel the least bit of guilt for being a stockholder, knowing how un-ethically Rambus has behaved as a company?

How were they unethical?

How is Micron, Infineon and Hynix using JEDEC to cherrypick patents when they were under NDA for Rambus technology "ethical"? Rambus has been cleared of any wrongdoing, and it's the INFRINGERS who are now being caught in their dirty deeds. Haven't you been reading?

Are you ignorant? In a commodity market, attaching an additional royalty payment rate, surely increases prices, and effectively sets a minimum base price for the item.

Particularly when the very people manufacturing this stuff are colluding to keep your prices high!

I'm not ignorant to answer your question, but I gotta ask - are you BLIND?

http://www.voluntarytrade.org/...00000094e10cd0901.html
http://www.pcworld.com/news/article/0,aid,116001,00.asp
http://www.internetnews.com/infra/article.php/3318861
http://www.eetimes.com/story/OEG20020621S0071

Right, yeah, that's all rambus's doing again, eh?

Collusion is illegal. What the memory manufacturers did to kill RDRAM was illegal. What they do to keep your DDR prices at a certain level RIGHT NOW is illegal. What Rambus and Intel did was PERFECTLY legal, since they were BOTH dependent on memory manufacturers to manufacture the stuff for them.

Instead, what they did was try to kill RDRAM via illegal collusion so they wouldn't have to pay a 2.5% royalty. They were greedy bastards who wanted all the money to themselves without paying a cent. So instead, they stole it via JEDEC.

Originally posted by: VirtualLarry
It wasn't expensive to mfg, at least not much more than other types of DRAM, AFAIK. It was the royalty/ransom payments to Rambus that made it expensive. Which is what would have made it hard to compete, price-wise, in the market against other royalty-free commodity memories. So that's why Rambus colluded with Intel to strong-arm the market, to effectively declare the need for the new Rambus memory by fiat, while at the same time, launching lawsuits against all of the other non-Rambus commodity memory mfgs.

The royalty rate for RDRAM was 1.5%. How was that a dealbreaker?

Rambus did not COLLUDE with Intel! To COLLUDE means to COMBINE EFFORTS TO DRIVE A COMPANY OUT OF COMPETING!

Remember, Rambus doesn't manufacture anything.
Intel doesn't manufacture RAM

They both NEEDED the cooperation of memory manufacturers. 7 of the 11 cooperated. Micron, Infineon and Hynix did not. They decided to KILL rdram by colluding to force a player out of the market.

So, how did they try to "strongarm" the industry? Hell, Intel paid Micron $500M to build RDRAM manufacturing processes, and they never produced a single module. Who "colluded" there? (By the way, Micron kept the money).

Originally posted by: VirtualLarry
I'm not going to claim that there wasn't some collusion in the industry to keep DRAM prices high, I think that many suspected as much, but what Rambus attempted to do to the industry is: 1) unrelated, and 2) much, much worse. And they're not "monkeys", they're engineers, and they are probably the ones least likely to engage in price-fixing collusion, that would be the marketing arms of those various DRAM mfgs if anyone.

1) It's not unrelated. Read the links above.
2) What did they DO that wouldn't have broken the LAW?

Originally posted by: VirtualLarry
No, I would never have thought that Rambus was responsible for any price-fixing going on between the non-Rambus DRAM mfgs. But it's interesting that you drag out that straw-man into the discussion, to attempt to make Rambus look good, by making non-Rambus mfgs look bad. Nice try. That doesn't make Rambus any less evil and unethical themselves.

You still haven't shown any clear pattern of behavior that makes Rambus look evil or unethical. Everyone's tried to show how Rambus is this evil company, yet no one looks at the cold hard facts. The cold hard facts is that Rambus did nothing illegal, while the three companies involved in these lawsuits DID.

Yet everyone's cheerleading for the ones who actually broke the law.

It just goes to show how a good smear campaign can make a piece of sh1t smell like a rose. That's exactly what you're doing by relying on the same old anti-rambus arguments that never held up in court.

Do some research. Familiarize yourself with the case. Read the dockets. Then come back and argue.

 

[VirtualLarry]

Originally posted by: Ice9
Read the '703. Read the court transcripts. Read the FTC docket. Read the Infineon dockets. Read the Markman ruling and learn the laws of Kingsdown instruction. Educate yourself, and see what both sides said. You won't do so though, because you, like many others, are far more happy hating Rambus when they shouldn't be the target of your dismissal.

Interesting legal buzzword quoting, when you apparently didn't even understand what the dismissal of summary judgement meant, assuming DerKaiser's comment about it is true. (If Hynix filed for a summary judgement against Rambus, asking for a dismissal, and the judge denied that motion for summary judgement, that DOES NOT mean that the judge decided that Hynix is infringing. Not at all. Only that a full court case should proceed, on the merits of the case. I haven't read anything about the case itself, but if that's what happened, then that's what it means, and you didn't understand the significance of the judgement at all, and are just reading into it like a fanboy.)

Originally posted by: Ice9
Anyway, your IP-only scenario is not what Rambus is. They are a technology company that develops intellectual property. They have no desire to manufacture or warehouse anything.

Which is exactly what an IP-only company IS.

A response like that means that I shall not participate further in this thread, because obviously some of it is flying over your head, and it would be pointless for me to continue. Enjoy your Rambus stock. While it's worth something.

Originally posted by: Ice9
And RDRAM, for the duration of its lifespan, proved to be a superior technology to SDRAM and DDR.

No, actually, it wasn't superior at all. In fact, that's one of the biggest reasons that Intel had to abandon their support of it. Add more RIMMs (for servers), and you increase the latency. The whole original benefit of RDRAM, was the reduced pin-count of the chip packaging, which was supposed to lower costs. Which it did in small, embedded applications (inside the N64 and PS2, for example), but because of the royalty rates, it wasn't cheaper than other DRAM types for desktops/servers, it was actually more expensive. Not to mention the thermal problems.

Originally posted by: Ice9
It wasn't until RDRAM was forced out of the market that DDR finally improved - albeit marginally.

RDRAM was forced out of the market because it just couldn't compete, even though it had a tiny overall speed advantage initially. Add the requirement for being able to add mass quantities of memory, and given the availability of dual-channel DDR systems - RDRAM became a "worthless" white-elephant memory technology, too hot and with too high a latency penalty. (It was a poor match for x86 systems, which tend to directly access memory in opcodes, (not being a load-store architecture like the MIPS CPUs, which were used in the N64 and PS2), and most PC apps have poor locality-of-reference. Rambus's only advantage was in bulk burst transfers, and multi-channel DRAM-banking schemes removed that advantage.

(In fact, if RDRAM was so good, then why isn't it currently being used in high-end video cards, instead of DDR/DDR2/GDDR3? Oh, right, because it's a poor fit for that application.)

Originally posted by: Ice9

Originally posted by: VirtualLarry
I wouldn't go so far as to say "encourage". More like it's allowed, legally, but that's about all.

No. It is encouraged. It is an accepted practice. It happens every day, it's a standard component of patent law. What is DISCOURAGED is to make these claims overly broad, which is not the case with Rambus.

"Submarine" patents are encouraged? Go tell an engineer that. He'll probably either laugh at you, or get angry at you. Of course if you talk about that to a greedy corporate patent lawyer, he'll probably start rubbing his hands in glee.

Originally posted by: Ice9
Do YOU understand what a patent is?

Yes, my father has a few engineering-related ones.

Originally posted by: Ice9

Originally posted by: VirtualLarry
That's completely BS. Rambus amended their existing patent filings, based on information that they obtained from the JEDEC forum, and never informed them (at the time) that they were filing additional patent claims based on that info, at least according to what I've read on the subject.

That is the case today. That was NOT the case when Rambus attended JEDEC.

Huh? Are you saying that they did or didn't do what I stated that I've heard that they alledgedly did? I don't understand how you can suggest that historical revisionism changed whether or not those actual acts took place, they either did or they didn't.

Originally posted by: Ice9
Now... Tell me about JEDEC's patent disclosure policy? The patent disclosure policy did NOT apply to patent applications at JEDEC, nor could it have since it was not legal to do so at the time.

Ok, "Mr. Lawyer", please point out where in the law it states that a 3rd-party group of technological standards collaborators, are legally barred from requiring or requesting that their member participants notify the group of any potentially-related patent filings?

Originally posted by: Ice9
So, if RAMBUS was not permitted by law to disclose patent applications,

Where the heck do you keep getting this from? It may be corporate policy, but I've never heard that it was prohibited by law to announce to the world that you've applied for a patent. In fact, I've heard in many interviews (unrelated to this issue), of people at companies that develop new tech, and when asked about it, they respond that they can't tell you about that tech, because they've applied for a patent on it. That's just so that some competitor doesn't steal the info and try to obtain a patent ahead of them. But they're in no way legally barred from disclosing to the world that a patent was applied for. I think that you are very confused.

Originally posted by: Ice9
and they were under no disclosure policy at JEDEC to disclose, and IBM, HP and Intel ALL declined to disclose THEIR patents (which was OK with JEDEC), where was Rambus in the wrong again?

I don't follow how IBM/HP/Intel fall into this, exactly.

Originally posted by: Ice9

Originally posted by: VirtualLarry
There's no law that says pending patent applications cannot be disclosed, in fact, it's often encouraged to announce. What's not encouraged, mainly to protect the patent-pending technology, is openly discussing the "secret details" contained within, basically telling how it works inside. But announcing that a patent has been filed, especially when you are participating in a standards-body forum, is the proper thing to do. Rambus withheld that information, instead (apparently) planning to hit up those very same companies some years later with lawsuits, for profit. (It is my opinion that they did this on purpose, because they knew that if the companies forming the standards, found out that the tech might have been patent-encumbered, then they might have changed the standard to work around that, and then Rambus wouldn't have been sitting-pretty a few years down the line, all set up to effectively legally extort an entire industry.)

So, you're saying that even though these companies KNEW about Rambus' technology, and KNOWINGLY voted it into the SDRAM standard having SEEN this technology under NDA, it was RAMBUS's fault for not saying anything about it when they weren't even legally obliged to do so?
That's pretty funny!

I'm not saying that at all, please don't put words into my mouth. I'm saying that you keep repeatedly asserting that it is outright illegal to announce a patent filing. I'm saying that I believe that you are wrong, based on my experience in general, and to try to use that false assertion to excuse Rambus's allegedly un-ethical behavior at JEDEC, is sad.

So enjoy your stock, you never know, you might run out of T.P. some day.. but this is my final post in this thread.

 

[DerKaiser]

Interesting legal buzzword quoting, when you apparently didn't even understand what the dismissal of summary judgement meant, assuming DerKaiser's comment about it is true. (If Hynix filed for a summary judgement against Rambus, asking for a dismissal, and the judge denied that motion for summary judgement, that DOES NOT mean that the judge decided that Hynix is infringing. Not at all. Only that a full court case should proceed, on the merits of the case. I haven't read anything about the case itself, but if that's what happened, then that's what it means, and you didn't understand the significance of the judgement at all, and are just reading into it like a fanboy.)

Sorry, I appear to be wrong on this one. The judge in the case ruled that Hynix did infringe 29 claims, but there are two phases of the trial to start. One in March and one in June. To get more info just go to Google and enter Rambus Hynix ruling.

I thought Hynix had not been found to infringe based on this link from January 5th of this year. Usually things are so slow to proceed in court that nothing significant could have happened in the meantime, but I was mistaken.

 

[VirtualLarry]

Originally posted by: Ice9

I thank the Lord above that Rambus wasn't involved with the W3C, and that I'm not paying a per-byte royalty charge to post this, an "IP tax" on IP protocol for instance.

If it meant your internet connection would be 3x faster for a marginally higher cost, why not?

Obviously, that illustrates your overall point-of-view pretty well, to everyone reading this. Of course you wouldn't care, because you would be the one taxing me (indirectly). But some of us prefer not to be taxed on the air we breath and the words we speak, thanks.

Originally posted by: Ice9
Rambus has been cleared of any wrongdoing, and it's the INFRINGERS who are now being caught in their dirty deeds. Haven't you been reading?
Particularly when the very people manufacturing this stuff are colluding to keep your prices high!
I'm not ignorant to answer your question, but I gotta ask - are you BLIND?

See the next response...

Originally posted by: Ice9
Instead, what they did was try to kill RDRAM via illegal collusion so they wouldn't have to pay a 2.5% royalty. They were greedy bastards who wanted all the money to themselves without paying a cent. So instead, they stole it via JEDEC.

I would just like to point out, that if you are on one hand arguing that the other memory mfg's colluded to fix prices (the price-fixing being the potentially-illegal act here, "collusion" is not in itself illegal), that price-fixing was to raise prices, to keep DRAM mfg's profit-margins high. Not lower them, as they would have had to do in order to drive a competing technology out of the market and kill it.

With that huge hole shot in your argument, I now bow out of this discussion.

(Not to mention, the fact that Rambus were greedy bastards, that wanted a percentage of the profits of DRAM sales worldwide all to themselves, without having to mfg a single chip, based on their abuse and manipulation of the standards-body process. So instead, they attempted to steal it, via the patent system.)

Originally posted by: Ice9
Yet everyone's cheerleading for the ones who actually broke the law.

There is often a bit of a difference between ethics and law. Sometimes the law allows stealing, in some contexts, even though it would be un-ethical to do so. What is alleged against Rambus is one of those cases.

Originally posted by: Ice9
Do some research. Familiarize yourself with the case. Read the dockets. Then come back and argue.

I don't need legal documents to help me decide what I know is wrong and what is wrong. Apparently, not everyone is blessed with that ability.

 

[Ice9]

Glad I waited to reply

Anyway, yes. The judge DID rule that Hynix infringed on 29 claims. There's never been a judgement of prior art, in fact the infringers never disputed the validity of Rambus' patents. They merely try to say that Rambus had a duty to disclose those patents during their tenure at JEDEC.

That defense falls flat on its face when you take into account the FTC ruling AND the prior CAFC ruling issued by Judge Rader of the federal circuit. They said there was a "staggering lack of defining details" in JEDEC's patent policy, which was way out of line when you consider every member sitting in JEDEC (the memory manufacturers anyway) are all in DIRECT COMPETITION with each other. IBM didn't reveal any patents despite having technology that read on the standard. HP as well. Micron as well. But rather than use those patents against each other, they colluded to squeeze out the little guy and smear their name in the press with the usual FUD that so many on this board have fallen for hook, line and sinker.

DerKaiser, i'm glad you did your due diligence on this one. You should read into it further, you'll be entertained at the very least, disgusted with the infringers and the memory fab industry at the very most.

I find it funny that this case parallels the "Intellectual Property" rights that everyone brings up with the RIAA - about how the little guys get squashed by the RIAA and their "intellectual property" rights and basically go bankrupt - despite quoting laws that protect them.

The same deal is happening to rambus. They are protected by the patent laws of the USA. But there's nothing to stop patent infringers from dragging things out in court until the little guy can't afford to fight anymore.

And for what? To stifle innovation and make more money with less.

 

[Ice9]

I would just like to point out, that if you are on one hand arguing that the other memory mfg's colluded to fix prices (the price-fixing being the potentially-illegal act here, "collusion" is not in itself illegal), that price-fixing was to raise prices, to keep DRAM mfg's profit-margins high. Not lower them, as they would have had to do in order to drive a competing technology out of the market and kill it.

With that huge hole shot in your argument, I now bow out of this discussion.

(Not to mention, the fact that Rambus were greedy bastards, that wanted a percentage of the profits of DRAM sales worldwide all to themselves, without having to mfg a single chip, based on their abuse and manipulation of the standards-body process. So instead, they attempted to steal it, via the patent system.)

Well, with this document effectively killing your argument, i'll stick around and see what other nonsensical tabloid trash gets thrown at this for me to debunk.

This is from Willie Meyer, Infineon's (then Siemens) head dram engineer who devised most of the plan to sink Rambus from the start. It has been part of evidence from the start:

http://www.ftc.gov/os/adjpro/d...rytradeamicusbrief.pdf

Search for the words "Deadly menace".

The ORIGINAL memo:

http://rambus.org/legal/menace.pdf

There is often a bit of a difference between ethics and law. Sometimes the law allows stealing, in some contexts, even though it would be un-ethical to do so. What is alleged against Rambus is one of those cases.

So, cheerlead the lawbreakers, but banish the people being ACCUSED of poor ethics?

Flawed logic indeed.

I don't need legal documents to help me decide what I know is wrong and what is wrong. Apparently, not everyone is blessed with that ability.

What you need is the whole story. You need the transcripts. You need to know what actually happened versus the hype and vitriol you read in trade rags. There's a real story here that you simply refuse to read.

I suggest you read it instead of burying your head in the sand.

 

[DerKaiser]

DerKaiser, i'm glad you did your due diligence on this one. You should read into it further, you'll be entertained at the very least, disgusted with the infringers and the memory fab industry at the very most.

Dude, you're too emotionally tied up with this thing. I couldn't really give a s**t one way or another.

Just because they have patents doesn't mean they acted ethically. And just because one judge says they infringed does not mean the matter is not debatable which was my original point. If you recall in the case of Rambus vs Infineon, the judge originally made a very narrow interpretation of the patent language. That was thrown out by an appeals court and the whole thing needed to be started over again. The same thing could happen here if Hynix decides to appeal.

To me it's pretty clear that Rambus amended their patents to cover as much of SDR and DDR RAM as they could regardless of whether or not they "invented" it.

 

[fishmonger12]

jedec: when monkeys collide

sounds like a good movie title

 

[grant2]

you guys are wasting your time arguing with ice9, he's already had this argument several times on anandtech and will beat you all with experience, whether he's right or wrong.

At least fkloster has given up his "rambus the company may die but rdram is here to stay..." mantra.

 

[Gamingphreek]

Originally posted by: grant2
you guys are wasting your time arguing with ice9, he's already had this argument several times on anandtech and will beat you all with experience, whether he's right or wrong.

At least fkloster has given up his "rambus the company may die but rdram is here to stay..." mantra.

I highly doubt that... Senior Forum member versus VirtualLarry a well respected and a Diamond Member. Yes titles and post count means nothing, however Larry is a very reputable forum member.

Correct me if i am wrong but isn't 1066 RDRAM still a little faster than DDR. THe only problem is latencies.

-Kevin

 

[Ice9]

Originally posted by: Gamingphreek

Originally posted by: grant2
you guys are wasting your time arguing with ice9, he's already had this argument several times on anandtech and will beat you all with experience, whether he's right or wrong.

At least fkloster has given up his "rambus the company may die but rdram is here to stay..." mantra.

I highly doubt that... Senior Forum member versus VirtualLarry a well respected and a Diamond Member. Yes titles and post count means nothing, however Larry is a very reputable forum member.

Correct me if i am wrong but isn't 1066 RDRAM still a little faster than DDR. THe only problem is latencies.

-Kevin

It's been a while, but it's likely. Latency has never even proven to be a problem in practice. The only time "latency" was a problem was back with the 820 chipset when there was only a single RDRAM channel on a 133mhz FSB. Not exactly making RDRAM shine.

The way I see it, RDRAM is old news. It was forced out of the market, resulting in the lawsuits we see today. The huge memory manufacturers that make billions a year knew the teeny Rambus company was a threat, and acted with whatever malice they could to squash 'em into obscurity. Rambus had the invention THEY needed to better themselves, so rather than license it, they used JEDEC to steal it.

Then when a BIG player called Intel came knocking, the DRAM industry realized "whoa, with Intel's backing, Rambus could be dictating what we do in our fabs" and again, acted accordingly.

I honestly don't care who has "diamond member" status or who is revered on the forums. People are misinformed about Rambus. There was a HUGE smear campaign that jaded the whole community against them, and they used the almighty dollar to do it. They price gouged RDRAM to $1200 per 128MB when the 820 was released. Did anyone truly believe a 128MB module cost that much to produce?

Well, I know the TRUE answer to that, but memory manufacturers would tell you otherwise. That is, until Mitsubishi and Samsung entered into evidence that RDRAM didn't cost much more to manufacture than SDRAM.

Do you honestly believe that with Intel backing RDRAM that there were "better technologies" available? Well, no, because they all infringed on Rambus's patents. The newest Hynix summary judgement kinda shows that.

Do you honestly believe that Michael Dell was wrong when he said DRAM manufacturers were engaging in "cartel-like behavior"? c'mon.

Look at where Rambus is today. They're in the forefront with RASER cells for PCIE. They have XDR, which nukes ANY dram technology out there today. They have Redwood interconnects, which CELL will use. They have Flexphase, which kills the trace-length limitation inherent in today's memory technology.

Rambus has matured as a company. They aren't perfect. They have made plenty of mistakes. But none as costly as the outright theft that memory manufacturers have made. All Rambus has to do is stay profitable so they can pay their legal bills. That means this little company has to STAY LEAN, keep engineering the solutions they have been, and stay 200% ahead of JEDEC.

And thus far, they've done so.

Love them or hate them, RAMBUS THE COMPANY is here to stay.