Originally posted by: razor2025
Huh? What kind logic is that? Please Ice9 check your logic before hittting the "submit" button. If it wasn't for the messed up IP / Patent system we have today, these stupid lawsuit wouldn't even have happened. Breaking the law doesn't neccesary equate with being unethical. Neither does using the law to your advantage can be considered ethical.
How so?
I'll tackle your post point by point. Please read carefully and completely.
Rambus developed the idea behind the RDRAM interface back in 1988. The same principles were applied to SDRAM during JEDEC in the mid-90's when they did their cherry-picking. Rambus was never allowed to present at JEDEC, but every company attending knew about RDRAM from NDA's dating back 5 years before Rambus ever attended... Where's the flaw in my logic?
When was Rambus supposed to reveal they owned this technology if they were barred from presenting?
Doesn't it appear odd to you that they would invite Rambus to be a member of JEDEC, yet NEVER allow them to present? It certainly seemed odd to the Judge at the FTC.
What RAMBUS is doing right now is completely unforgivable. Instead of striking out profitable liscencing deals (Next-Gen consoles), they would rather sink enomous amount of company resource into hitting the "sue-happy" lottery for loyalties from 5 years ago. Any company that base major or entire revenue off of lawsuits are POS in my book. Look up SCO.
It's unforgivable to want to be PAID for your invention that was blatantly stolen from you?
There is no parallel here. SCO is attempting to claim that portions of UNIX code are present in Linux. They (SCO) have so far not been able to prove this in a court of law beyond a reasonable doubt.
Rambus HAS been able to prove infringement, as they have just been awarded summary judgement in the hynix case. The companies who are infringing on their patents ADMIT they are infringing. Where's the flaw in my logic? What's in question here is whether or not Rambus took advantage of their position in JEDEC or whether JEDEC took advantage of having RAMBUS being a member and never letting them present.
The bottom line here is that everything Rambus did during their tenure in JEDEC was 100% legal. What the members of JEDEC did was NOT legal (collusion, price fixing).
Like I said before, Rambus made plenty of mistakes. But they never broke the law, and they wanted to work with these companies to manufacture their product. Infineon, Hynix and Micron however DID break the law in an effort to kill Rambus as a company.
Also, because I'm a consumer, I don't see how RAMBUS winning the case will benefit me. RAMBUS stock owners might get a hefty profit when RAMBUS wins, but it's crap deal for consumer. If RAM factories has to pay loyalties to RAMBUS, they'll be sure to jack up the price to compensate for that. But I guess, it's pointless to argue against someone who has vested interest in such a POS company.
Rambus RDRAM royalties were 1.5% of manufacturing cost, which essentially makes your cost $1 or $2 more. RDRAM was superior and less costly to manufacture than SDRAM (samsung and mitsubishi both testified to that fact in the FTC trial and the virginia trials). It was SO much faster than SDRAM, it makes you wonder why JEDEC even bothered. If RDRAM was cheaper to manufacture and was so much better than SDRAM, wouldn't that $2 premium be worth it to you as a consumer? Are you sure it's my logic that's flawed here?
What would you rather do, pay $9,000 for a Yugo or $10,000 for a ferarri?
And this whole RAMBUS - JEDEC relationship. It's stupid and naive. They knew that RAM factories were bunch of Mafia thugs, but they seemed to be naive enough to think that they might get to cut a liscencing deal if they showed their stuff at JEDEC. What kind of dumb@$$ company will allow bunch of other company to see their IP in a forum like meeting? Especially, when all the members are mafia-like entity? It's like RAMBUS wanted their IP to be stolen.
No. It's not like that at all. Rambus doesn't manufacture anything. They are a company of engineers. They want memory manufacturers to manufacture the stuff and bring it to market as part of a strategic partnership. This is why they shared their innovations under NDA in the first place (in 1990). They are a tiny company that has brilliant ideas, but no cash for fabrication plants. They simply didn't have the resources to manufacture mass quantities of RDRAM, XDR or any other type of chip. They supply the brains, the memory manufacturers supply everything else.
They were invited to JEDEC with the premise that they would be allowed to present their technology for standardization (meaning RDRAM would have been the standard). JEDEC rules stated that they are willing to license technology deemed as "proprietary" and use it as part of their standard. Rambus was there to present their technology to JEDEC for consideration. However they were barred from presenting their technology.
Had Rambus been allowed to present their technology, the memory manufacturers would not have been able to try and steal it, as it would have put them "on notice" that they were infringing on someone else's technology. At the time, Rambus thought they were there for the purpose of presenting. They found out soon enough that it wasn't why they were invited.
Read the FTC docket on this. This is why Rambus won there.
Also, RDRAM FAILED because it could not compete against SDR/DDR during the time it was unveiled.
No. RDRAM failed because the memory manufacturers lied about production schedules, production costs and demand. Memory manufacturers said to Intel there was no demand for the stuff. How could that have been if Dell had to charge $1200 per 128MB module if you wanted a 256MB system?
This is currently the topic of investigation of a Department of Justice probe. Michael Dell himself complained about these memory manufacturer tactics and said they were engaging in "Cartel-like behavior". You may have heard about several Micron executives being subpoenad and several Infineon executives going to jail in connection with this.
Oh, and one other thing. RDRAM hasn't "failed". It's still in plenty of products being manufactured today (HDTV's, PS2's, high-speed routers, fiber multiplexors, several other products that demand high speed and low pin counts).
Add the fact that AMD was smart enough to keep their Athlon platform in SDR and DDR. These two things brought down the viability of RDRAM as a product.
At one point, AMD was a Rambus licensee. However, at the time they had single-digit market share, and less than 15% market share now. AMD might be regarded as some to be a "major player", but the sad truth is that as good as AMD products may be, they cannot compete with Intel's marketing muscle & corporate ties. This is why having intel on your side is a far better thing than having AMD on your side.
Intel intended to scale their P4 architecture into very high FSB/Clock. So, on paper, RDRAM looked like a good RAM technology to pair up with their new NetBurst (netcrap) architecture.
Bash Netburst and Intel all you like, they own 80% of the desktop market to this day. And RDRAM *WAS* a good technology to pair up with the P4. It was a good technology for a great number of things. It's still in use today for high bandwidth applications (like HDTV's). It would have been a lot more beneficial if it were given the chance. But the memory manufacturers colluded to keep it out of the market. Again, this is the topic of a DOJ probe.
However, due to Intel's overzealous marketing and craptacular CPU/Chipsets, their older generation (P3) and AMD's newest Athlons b*tch slapped early P4. If RDRAM was still around and supported by Intel today, they might be a good match against DDR due to the high FSB achieved by P4 currently. But.. that's not the case. RAMBUS was too arrogant to let RDRAM mature and wait till Intel had a solid platform to utilize their RAM.
The only "immature" chipset released for the RDRAM platform was the i820. A single-channel 133mhz front side bus is simply not going to make an 800mhz dram look incredible. RDRAM had already matured in the market as it was in a previous product (the Nintendo 64), which SGI chose as their graphics memory. It proved itself there, and that was what turned Intel in their direction in the first place. i820 was more a "proof of concept" thing, but it wasn't until the P4 showed itself that RDRAM proved itself as a viable PC technology.
Nope, they had to be greedy and rode the Intel wagon. They thought that by using Intel's enormous marketing machine and its (practically) monopoly status, they could've been the RAM technology top-dog. Unfortunately, things didn't work out like they planned, so Intel dumped RDRAM like a used whore. And we aren't even talking about the cost of RDRAM ownership at the time. Even without considering the cost of RDRAM, their performance alone sucked major.
No. It was Intel who approached Rambus, not the other way around. Intel said they had an interest in their technology to mate with the Pentium 4, with a test run in the late-model Pentium 3. Again, read the court transcripts when you have the time and want to know the truth.
The performance of RDRAM on the P4 was better than anything out at the time. i845 for both SDRAM and DDR never caught up speed-wise, and Granite Bay still couldn't keep up. It wasn't until DDR scaled to 400mhz years later that RDRAM was finally bested by DDR. And even then, PC1200 RDRAM killed it even though Grantsdale was never released to the public. SiS promised a chipset for PC1200, but again - engineering samples made it to reviewers but SiS dropped it before it could see the light of day (no doubt thanks to the memory manufacturers who colluded to keep IT out of the market too).
Intel made the right decision regarding discontinuing RDRAM support. Memory manufacturers had effectively squeezed it out of the market (again, the source of a DOJ probe into their anticompetitive behavior). Rambus had always contended that this was going on, and now they have evidence to prove it thanks to Mitsubishi, and now (surprise!) Infineon.
But from intel's point of view, they couldn't sell CPU's if memory manufacturers wouldn't manufacture RDRAM. Thus, the 845 was born to support SDRAM, which memory manufacturers WERE manufacturing in quantity.
One major thing that stands out though. WHY did RAMBUS not sued the RAM factories in the very early 1990?
Simple. Rambus doesn't manufacture anything. They NEEDED the DRAM manufacturers to bring their product to market. Also, in 1990, there was no such thing as SDRAM. It was all EDO SIMMs, and intel was complaining that memory speeds weren't fast enough to keep up with CPU scalability. It wasn't until 1992 that Rambus was invited to JEDEC, and it took them some time to figure out exactly what was going on. Once they found out that their patents were being cherry-picked one by one, they did the only thing they could do - amend their patent claims to cover SDRAM so their intellectual property would be protected from theft.
While Rambus did this, they did everything they could to get JEDEC to allow them to present their technology for consideration. To this day Rambus is the only company in the history of JEDEC to have been barred from presenting their technology. They were never given a clear reason as to why. In 1996, Rambus gave up and left JEDEC stating that they would persue any avenue to protect their intellectual property.
If they knew that their had their IP stolen after the 1990 JEDEC, why haven't they filed suit immediately afterwards? Or even few years after? Instead, they waited till 8-10 years later, and brought out the lawsuits hoping to catch some loyalty lottery that they were "owed" since then.
This is the correct timeline:
1988 - Rambus invents RDRAM
1990 - Rambus files patents for RDRAM and shares their technology model with Memory Manufactures in hopes of bringing it to market. This is done under NDA.
1992 - Rambus invited to JEDEC (this is shortly after Willie Meyer of Infineon
wrote the "Deadly Menace" memo and sent it to Gordon Kelly, chairperson of JEDEC)
1995 - JEDEC finalizes the SDRAM standard
1996 - Rambus patents are GRANTED by the USPTO and subsequently leaves JEDEC.
The lawsuits did not start until 2000-2001.
At the time, there were 11 memory manufacturers that were manufacturing SDRAM. They all received letters stating that they were infringing on Rambus' original 1990 patent filing. Of these 11 companies, 8 of them agreed to pay royalties without much fanfare (Hitachi gave a little fight because they wanted better terms, but ultimately paid up). The last three (Infineon, Micron, Hynix) decided to go right on infringing.
Hope this explains things for you.