Hender if you knew all the BS corporate scheems that all sides do than you wouldnt be buying much of anything. Its just buisness, as long as they arent beating kids or killing whales its just part of the game. Infact most of the clothes you have on are probably made in sweatshops and most of the fruit you eat picked by migrent workers payed slave wages with children ages of 12 out in the sun but you buy them anyways. I would be more offended by the latter than some corporate hoohaa that you think Rambus alone has engaged in. Such ignorance
This is laughable to say the least. First of all Rambus isn't a retailer. They're an intellectual property company. They're a tiny company with 140 engineering geeks, 36 tech writers and clerical staff, and 4 lawyers. That's it.
The fact is, they've developed high bandwidth solutions that more than quadruple JEDEC's own "inventions" (which, SURPRISE, Rambus owns). The fact that they've single-handedly brought the memory industry to its knees shows just how weak a foundation JEDEC is. Nothing ever gets done by committee, and the one thing that SHOULDN'T is memory technology. If it weren't for JEDEC trying to keep RDRAM prices artificially high and DUMPING DDR ON CONSUMERS to keep it "popular", we wouldn't be stuck at the paltry frontside bus speeds we're using today.
As far as their "business" practices, people crack me up here
They think that corporate espionage and sicking the FTC on threatening competition is a good business practice Anyway the LAW says Rambus is not guilty of the fraud that everyone seems to love hearing about, and therefore have every right to sue for royalties on DDR and SDRAM. So, "shady" or not, the LAW says they have a right to royalties, and rambus will persue them. In the end Infineon has failed to beat Rambus in court. In order to say that their business practices are shady, you have to feel that intellectual property has no place in technology.
And I'll put my faith in Rambus to deliver a speedy memory technology for 12ghz cpu's that Intel has on their roadmaps over some slothful snail-paced committee like JEDEC.
No flames, and yeah... Wait for the benchmarks.
And you're right, the general public won't care about it
Only performance enthusiasts will. What I want to see is Intel's reaction to it if it's a faster solution than their own chipsets. Intel has NEVER liked anyone taking the performance crown away from them. They've ALWAYS responded with something better. And something tells me that it won't be something modeled after Granite Bay
Rambus will be cheaper than DDR once a few of those pesky DDR companies go out of business (Micron, Hynix and Infineon are all WELL on their way, in fact Micron isn't projected to last the year). DDR companies have been dumping it for so long at an enormous loss, there's no way that technology could survive on a financial level, much less a technological one.
The only real reason the average tech guy "loves" DDR is because it's *cheap*, and they don't give a rat's ass if it pushes people out of work as a result.
I have never owned a Sis chipset and have not heard anything about Sis' reliablity as of lately but I do own 4-16bit 256mb PC-800 samsung modules and just like I would do to a sexy x-girlfirend that broke up w/me 6 years ago, I would 'stick' them in this new Sis-659 chipset and see what happens...
Cerb
Elite Member
- Aug 26, 2000
- 17,484
- 33
- 86
The price is skewed for the mobo/chipset, as one is new and one is old (er, matured
). It likely won't take that long to go down to $180 or so (though both are still too expensive, when a good NF2 board can be gotten for $120), making it even. In addition, you save $20 by going with Kingston instead of Samsung on the DDR ($68 vs. $79), bringing it down to $366, and the Gigabyte GA-8SQ800 is only $112, cutting it down to $248. Lies, damn lies and statistics, anyone?Also, the GB boards are going down, you can almost get one for $150 now.
Rambus is more expensive based on the memory, regardless of chipset, but the lawyers are the problem.
The biggest problems with Rambus right now are (a) consumer opinions, (b) DDR being pushed partly due to that, and due to the prices tending to be lower, or at least there are some that are (there's always a couple nice deals).
On the PC side, Rambus may not be going away, but for now, it is dwindling (the latency really hurts it--that's "really hurts" kinda like the dual-DDR of Nforce boards "really helps"--in a few years that should be no big deal, as it will be far faster).
On the server side, Rambus hasn't stopped at all, and the 1200Mhz RDRAM just begs for dual Xeons, and might allow them to wipe the floor with AMD duallies (mainly because both are memory-crippled now, but the AMD loses less from it, and the Xeons should gain more from more bandwidth).
Saying that something is "just business" is a way of shirking responsibility. It's like kids who get in fights and beat the snot out of some other poor kid. "Boys will be boys." That's crap. Someone has to take responsibility for actions. Rambus may not have violated the law in the courts' eyes, but what they did was sneaky, underhanded, and a violation of both the trust of the body that establishes memory standards and an abuse of the patent system that we all now is flawed. This isn't a discussion about sweatshops, either, but about Rambus. Try and stay on topic.
This is laughable to say the least. First of all Rambus isn't a retailer. They're an intellectual property company. They're a tiny company with 140 engineering geeks, 36 tech writers and clerical staff, and 4 lawyers. That's it.
The fact is, they've developed high bandwidth solutions that more than quadruple JEDEC's own "inventions" (which, SURPRISE, Rambus owns). The fact that they've single-handedly brought the memory industry to its knees shows just how weak a foundation JEDEC is. Nothing ever gets done by committee, and the one thing that SHOULDN'T is memory technology. If it weren't for JEDEC trying to keep RDRAM prices artificially high and DUMPING DDR ON CONSUMERS to keep it "popular", we wouldn't be stuck at the paltry frontside bus speeds we're using today.
As far as their "business" practices, people crack me up here
Anyway the LAW says Rambus is not guilty of the fraud that everyone seems to love hearing about, and therefore have every right to sue for royalties on DDR and SDRAM. So, "shady" or not, the LAW says they have a right to royalties, and rambus will persue them. In the end Infineon has failed to beat Rambus in court. In order to say that their business practices are shady, you have to feel that intellectual property has no place in technology.
And I'll put my faith in Rambus to deliver a speedy memory technology for 12ghz cpu's that Intel has on their roadmaps over some slothful snail-paced committee like JEDEC.
Yes, I realize that they're an intellectual property company. OTHER PEOPLE'S intellectual property, in some cases.
Rambus didn't break the law in the eys of the courts, but as I said in an earlier post, they did something very sneaky and underhanded. Yes, it happens a lot in the world of business. The vast majority of companies do things that are reprehensible, but what I dislike so much about Rambus is that they did it so plainly and everyone coudl see what they did--which is why I draw attention to it now. They sat in on the JEDEC meetings, listened to the proposed specs for next generation RAM that other people developed, then they went and patented those technologies. They abused their position on the board that establishes standards. So what if it moves slowly? They could have helped to speed things up, but instead they did something that I *know* is illegal, even though they won on appeal. What would you think if a company that sits in on the World Wide Web Consortium went and patented a technology that they had no position in developing, just that they were the first one to do it? Any web browser that wanted to render the new standard would be forced to pay royalties somehow.
The LAW isn't always right. I don't want to get into a legal debate, but they exploited a loophole in the system and won. They got off on a technicality, essentially. They're the business equivalent of the murderer who gets off because the police didn't collect the weapon properly or didn't fully read him his rights. It's was a crap decision, and they're profiting from other people's work. They patented work designed to be part of a standard, not belonging to a specific company.
That's shady and underhanded, and I refuse to buy products designed by them. That's my choice and I don't understand why I'm getting ragged on so much.
But these boards still can't beat RDRAM in real world benchmarks, and only come close in synthetic ones... PC1066 is what, 2 years old now? What took DDR so long to get this far? Could it be... oh, I don't know... manufacturing costs for 15 million layer PCB boards to support 256 pin-traces that all have to be an identical length?
The lawyers are not the problem. The problem is memory manufacturing companies who think they can make a company "just go away" by dumping a competing product below manufacturing cost. Dumping is *illegal*, by the way
RDRAM is *artificially* expensive, and DDR is *artificially* cheap. The only company making money on memory these days is... Samsung (who is Rambus' best friend).
Take a look at all the DDR companies like Micron, Infineon and Hynix. Micron is nearly dead (they just had to take a 500M loan in the form of convertible notes just to stay in business another quarter). Hynix is taking subsidies from the korean government to stay afloat. Infineon, well, their parent company isn't all too happy about their large losses either. The only saving grace for Infineon is that they have the luxury of being able to get out of the memory industry completely if they had to.
Which are almost always wrong...
Because no one would buy it if it cost as much as RDRAM... So instead, they take a loss and hope Rambus will just go away under the weight of the courts.... which SURPRISE! isn't happening
If this "latency" hurts so much, why is PC1066 RDRAM still the fastest memory technology for the P4 (proven TIME AFTER TIME), besting dual channel DDR which requires 8 TIMES the pincount and nearly DOUBLE the manufacturing cost in the form of additional PCB layers?
There is no RDRAM on the server side. Never has been. Memory bandwidth is NOT critical on servers like most people think. The average high-traffic server is limited by the network interface, NOT its memory throughput. This is why it's only now that big IBM and Compaq servers are shipping with DDR, and even so only in the PC2100 flavor.
When high memory throughput is demanded, it's nearly ALWAYS RDRAM now, and not DDR. The last 2 high-powered systems that actually made it to sale is the Cray X1 and the HP Alpha EV7. These machines go to research facilities where insanely fast speeds are required. DDR will never tread here.
Oh, and for any of you who "hate" Rambus, it's ok to hate a company, but be right about it. And make sure the company you "love" isn't guilty of FAR WORSE STUFF.. Read that and tell me who needs a class in proper business etiqutte.
They mentioned that this chipset is designed for the professional environment, if that wasn't obvious already. I think that it may in fact boost performance somewhere around 50% in certain professional applications. I don't think it will give solid 50% boosts all around, but in a few applications it might.
Sorry. I gotta take this one on point by point. The average person's hate for Rambus is often VERY misplaced (and in some cases flat out wrong).
Keep one thing in mind when reading this. I've been following the Rambus v. JEDEC trials/legal proceedings VERY closely for YEARS now. I have read just about every document there is to read on the subject. I can provide you with links to just about anything I state in any of my posts.
Most of all, try to understand that I support Rambus because I truly believe they are the REAL innovative force behind the memory industry, and they should be recognized as such. They've proven themselves to be the smallest, MOST EFFICIENT memory technology developer of our time. If you refuse to recognize that, then just give up reading anything I post.
Now, on with the show:
Not according to the CAFC.
Again, wrong.
This isn't what happened at all. Read the entire court text. I've been following these cases for years.
What happened was:
1. Rambus was INVITED to JEDEC. They didn't seek membership.
2. JEDEC sat in on committee meetings. This made several JEDEC members nervous, particularly Micron.
3. They realized that they ALREADY owned the patents to the signaling technologies they were discussing (FACT: These meetings happened in 1992, Rambus filed for these patents in 1990.)
4. Realizing that they had a potential gold mine, they quietly left JEDEC... and this is an important point:
THEY LEFT JEDEC BEFORE ANY SDRAM TECHNOLOGIES WERE RATIFIED AS A STANDARD. Hey, if it were MY standard, and I did the engineering and research into it, I'd have done the same thing. I'm in business to make money, not satisfy an open standards committee with free engineering.
Now, fast forward to the initial Rambus vs. Infineon trial...
Rambus' entire case hinged on one thing: The definition of a bus, since this is what was in dispute with regards to SDRAM's signaling technology. Rambus had filed for the patent in 1990, well before this lawsuit (or even JEDEC itself) was relevant. The initial "Markman ruling" ruled in Rambus' favor, but SOMEHOW, infineon's lawyers managed to get Judge Payne to THROW OUT the markman rulilng that defined Rambus' definition of a bus.
This means that Rambus had no way to define their technology at trial. The fact that they innovated this bus technology back in 1990 had no bearing, and the jury couldn't hear it because the judge ordered the Markman ruling as inadmissible. In effect, Rambus was ROBBED.
Watch a few Legal dramas on TV. You'll realize that a Lawyer's biggest job when defending a guilty client is to SUPPRESS VALID EVIDENCE, and KEEP IT FROM THE JURY. This is EXACTLY what Infineon's lawyers did to Rambus when they got Judge Payne to overturn the markman ruling that defined Rambus' own definition of a bus back in 1990.
So here's the BS reasons why Rambus lost their trial and were convicted of fraud:
1. Rambus couldn't use their definition of a bus (which was established in 1990 via a patent application)
2. Because they couldn't use their idea of a "bus", they had no way to defend against Infineon's claim of sitting in JEDEC meetings and patenting technologies (technologies it *ALREADY OWNED!*)
3. All the jury could hear was that Rambus sat in JEDEC meetings and wrote up patents to that effect.
The trial all boiled down to who had the better lawyers. Hell, infineon had Ken Starr. Remember him? The guy who investigated Clinton when he took that BJ from ole Monica?
I'm personally HAPPY that the CAFC recognized this gross error and overturned the ruling 100%. Again, this was NO small ruling. It was a 100 percent reversal of the lower court's ruling. Not a partial reversal. A full 100% reversal.
No, they LEFT JEDEC so they WOULDN'T do so, on the advice of their own lawyers. Rambus had NO WAY of knowing what JEDEC would ultimate ratify since that didn't happen until well AFTER rambus left JEDEC. But hey. Since Rambus is loaded with engineers, they were certainly capable of making a VERY educated guess. This made Rambus the SMARTER company.
Then the W3 consortium would have to PROVE in COURT that the patents weren't pre-existing, possibly involving some legal maneuver of the markman ruling I outlined before.
Better question: Would the w3 consortium be STUPID enough to invite an intellectual property firm to an open standards committee? JEDEC was
Hmmmm. Filed for patent in 1990. SDRAM standards ratified years later. Does this loophole involve Time Travel?
What technicality is this?
SILLY!
Becaue your assumption is wrong, that's all.
Oh, one other thing that you probably won't like one bit: If you purchase memory made by anyone other than Infineon, Micron or Hynix, then you're ALREADY paying Rambus royalties. Samsung, Elpida, Toshiba, Hitachi and the rest of the big memory manufacturers ALREADY recognize Rambus as the legitimate owners of SDRAM and DDR and pay them a 3% royalty on every chip manufactured. Sorry!
If you want to support DDR and NOT support Rambus, then you will have to limit your memory purchases to one of those 3 failing memory manufacturers :0
Keep one thing in mind when reading this. I've been following the Rambus v. JEDEC trials/legal proceedings VERY closely for YEARS now. I have read just about every document there is to read on the subject. I can provide you with links to just about anything I state in any of my posts.
Most of all, try to understand that I support Rambus because I truly believe they are the REAL innovative force behind the memory industry, and they should be recognized as such. They've proven themselves to be the smallest, MOST EFFICIENT memory technology developer of our time. If you refuse to recognize that, then just give up reading anything I post.
Now, on with the show:
Not according to the CAFC.
Again, wrong.
This isn't what happened at all. Read the entire court text. I've been following these cases for years.
What happened was:
1. Rambus was INVITED to JEDEC. They didn't seek membership.
2. JEDEC sat in on committee meetings. This made several JEDEC members nervous, particularly Micron.
3. They realized that they ALREADY owned the patents to the signaling technologies they were discussing (FACT: These meetings happened in 1992, Rambus filed for these patents in 1990.)
4. Realizing that they had a potential gold mine, they quietly left JEDEC... and this is an important point:
THEY LEFT JEDEC BEFORE ANY SDRAM TECHNOLOGIES WERE RATIFIED AS A STANDARD. Hey, if it were MY standard, and I did the engineering and research into it, I'd have done the same thing. I'm in business to make money, not satisfy an open standards committee with free engineering.
Now, fast forward to the initial Rambus vs. Infineon trial...
Rambus' entire case hinged on one thing: The definition of a bus, since this is what was in dispute with regards to SDRAM's signaling technology. Rambus had filed for the patent in 1990, well before this lawsuit (or even JEDEC itself) was relevant. The initial "Markman ruling" ruled in Rambus' favor, but SOMEHOW, infineon's lawyers managed to get Judge Payne to THROW OUT the markman rulilng that defined Rambus' definition of a bus.
This means that Rambus had no way to define their technology at trial. The fact that they innovated this bus technology back in 1990 had no bearing, and the jury couldn't hear it because the judge ordered the Markman ruling as inadmissible. In effect, Rambus was ROBBED.
Watch a few Legal dramas on TV. You'll realize that a Lawyer's biggest job when defending a guilty client is to SUPPRESS VALID EVIDENCE, and KEEP IT FROM THE JURY. This is EXACTLY what Infineon's lawyers did to Rambus when they got Judge Payne to overturn the markman ruling that defined Rambus' own definition of a bus back in 1990.
So here's the BS reasons why Rambus lost their trial and were convicted of fraud:
1. Rambus couldn't use their definition of a bus (which was established in 1990 via a patent application)
2. Because they couldn't use their idea of a "bus", they had no way to defend against Infineon's claim of sitting in JEDEC meetings and patenting technologies (technologies it *ALREADY OWNED!*)
3. All the jury could hear was that Rambus sat in JEDEC meetings and wrote up patents to that effect.
The trial all boiled down to who had the better lawyers. Hell, infineon had Ken Starr. Remember him? The guy who investigated Clinton when he took that BJ from ole Monica?
I'm personally HAPPY that the CAFC recognized this gross error and overturned the ruling 100%. Again, this was NO small ruling. It was a 100 percent reversal of the lower court's ruling. Not a partial reversal. A full 100% reversal.
No, they LEFT JEDEC so they WOULDN'T do so, on the advice of their own lawyers. Rambus had NO WAY of knowing what JEDEC would ultimate ratify since that didn't happen until well AFTER rambus left JEDEC. But hey. Since Rambus is loaded with engineers, they were certainly capable of making a VERY educated guess. This made Rambus the SMARTER company.
Then the W3 consortium would have to PROVE in COURT that the patents weren't pre-existing, possibly involving some legal maneuver of the markman ruling I outlined before.
Better question: Would the w3 consortium be STUPID enough to invite an intellectual property firm to an open standards committee? JEDEC was
Hmmmm. Filed for patent in 1990. SDRAM standards ratified years later. Does this loophole involve Time Travel?
What technicality is this?
SILLY!
Becaue your assumption is wrong, that's all.
Oh, one other thing that you probably won't like one bit: If you purchase memory made by anyone other than Infineon, Micron or Hynix, then you're ALREADY paying Rambus royalties. Samsung, Elpida, Toshiba, Hitachi and the rest of the big memory manufacturers ALREADY recognize Rambus as the legitimate owners of SDRAM and DDR and pay them a 3% royalty on every chip manufactured. Sorry!
If you want to support DDR and NOT support Rambus, then you will have to limit your memory purchases to one of those 3 failing memory manufacturers :0
I was simply pointing out your hypocrisy. You wont buy from Rambus due to their "shady practices" yet you admittably hide behind your own ignorance. "it happens a lot in the world of business. The vast majority of companies do things that are reprehensible, but what I dislike so much about Rambus is that they did it so plainly and everyone coudl see what they did" In other words its ok because I dont know I only dislike because im told to dislike not because they are dislikeable but because others would agree with me. BAAAAAAAAAA goes the sheep. The truth is they both ripped off each other. The JEDEC was only able to create DDR due to Rambus's patents in the early 90's while Rambus was only able to create RDRAM while being part of the JEDEC. So technically since I now told you they are both shady I guess you wont be buying any memory at all will you, so goes the righteous fool. Its just dumb, I wont buy Windows because they ripped off Apple, ya lets see how far that will get you.
Actually... Rambus had the signaling/bus patents for "sdram" filed for in 1990, at least 2 years before JEDEC was even relevant. JEDEC was the one trying to exploit it, since Rambus clearly had the engineering skill. This is why they were invited to JEDEC despite heavy reluctance from many JEDEC partners. Most of JEDEC's partners thought of (and still think of) Rambus as a threat because they are capable of engineering technology far more quickly than a large committee.
This goes against JEDEC's philosophy of "hanging on to old memory fabs for as long as we can to maximize profits while keeping memory interface speeds at the lowest acceptable level possible." The funny thing is how badly this backfired on them
Most people who bash Rambus don't know the truth, even though it's all readily available if you find REPUTABLE sources (READ: sources that don't have "crucial" aka "micron" banner ads littering their front pages.)
ed21x
Diamond Member
- Oct 12, 2001
- 5,411
- 8
- 81
Despite all the antagonism to Rambus' business practices, there's no denying that upon initial release, they have always had the best performance, and in the end, isn't that all that matters? I'm expecting this new type of memory to involve 32-bit memory installed in pairs, which really isn't too bad at all (or too expensive).
64-bit RIMMs are slated for Late '03 on Samsung's roadmap (link in one of my previous replies), I would expect it to coincide with this chipset release. But who knows?
dullard
Elite Member
- May 21, 2001
- 26,031
- 4,659
- 126
Ice9, nice posts. I've tried the same argument on other Rambus bashing threads. I just cannot convince anyone - so I've stopped trying. If you file a patent, then later discuss something in a committee, you still own everything in the patent. Seems like a simple idea, but people just cannot grasp it. Of course anything not covered by the patent that a company develops with the committee doesn't belong solely to the company. But that isn't the case with Rambus.
I'm in the exact same position as Rambus. My research into developing a DNA amplification machine resulted in several patents. In an attempt to gain more research cash, we signed an agreement with another company. That agreement said the company will fund more research and then they have the first rights to sell the finished product (with paying us royalties of course). Now last week they failed to meet the 2nd agreed payment and had their lawyers send us a letter notifying us of some changes to the contract that they would like to make: (1) they will no longer make any payments, (2) they get 12% of anything we make off this machine. Why on earth would they have the right to get 12% of profits from a machine that we developed on our own with our own money, we got the patents on the machine before contacting you, and you won't obey the contract that we signed? I guess I belong to a ruthless company since I think that behavior of the 2nd company is laughable - now everyone will boycott my product.
I'm in the exact same position as Rambus. My research into developing a DNA amplification machine resulted in several patents. In an attempt to gain more research cash, we signed an agreement with another company. That agreement said the company will fund more research and then they have the first rights to sell the finished product (with paying us royalties of course). Now last week they failed to meet the 2nd agreed payment and had their lawyers send us a letter notifying us of some changes to the contract that they would like to make: (1) they will no longer make any payments, (2) they get 12% of anything we make off this machine. Why on earth would they have the right to get 12% of profits from a machine that we developed on our own with our own money, we got the patents on the machine before contacting you, and you won't obey the contract that we signed? I guess I belong to a ruthless company since I think that behavior of the 2nd company is laughable - now everyone will boycott my product
.Dullard, I know what you mean. I dealt with this kind of ignorance over at HardOCP.
I have to admit though, it's funny to watch Rambus develop this sort of technology and announce these amazing memory technology products, only to have people defend the slower, more inferior stuff.
It's just a case of people supporting what they buy, even if they're hurting the company bringing it to them.
I think what surprises me the most on these forums is the amount of LOVE they have for the little guy (AMD vs. Intel, for instance) in EVERY OTHER REGARD. Normally people in here LOVE a story where David slays Goliath. Now that it's happening, it's funny to watch people cry foul
But now, you find these DDR zealots praising their multi-billion dollar memory manufacturing conglomerates (Micron, Infineon, Hynix) when it's the little 180 person company that has the potential to be screwed out of developing a better product.
I thought i read something a little while back that pointed out that Rambus did issue the patents in 90 but they were vague. They took some ideas they got while at the JEDEC and incorperated them with there patents to make RDRAM complete. But at the same time the JEDEC took the ideas that Rambus had patented when Rambus joined them and implemented them to make SDRAM and now DDR. So i think they store a little information technology from each other to make there own ram, one is no better than other. But honestly I dont give a damn about who has the bigger halo over there head, as long as neither one does anything atroshious (like I said before) than it really shouldnt be an issue. I figure that its just an excuse for the majority of the people to feel good about there purchase.
But that source might have been wrong, I havent really read extensivly on it because I just dont care.
But that source might have been wrong, I havent really read extensivly on it because I just dont care.
lol while reading the other forum i felt like jumping in and arguing with them
Wish i could meet all these people face to face and just argue up a storm
DAMNIT MORE IM READING THE OTHER FORUM MORE IM GETTING UPSET, SUCH IGNORANCE, SUCH STUPIDITY AND SUCH A BROAD BIAS TWARDS DDR THAT MOST WOULD THINK ITS THE NORM
Wish i could meet all these people face to face and just argue up a storm
DAMNIT MORE IM READING THE OTHER FORUM MORE IM GETTING UPSET, SUCH IGNORANCE, SUCH STUPIDITY AND SUCH A BROAD BIAS TWARDS DDR THAT MOST WOULD THINK ITS THE NORM
lol, I think the 3 of you (dullard, Ice9, & DXplayer) scared off all the rambus bashers!!! Its an Anandtech FIRST!!!! Nicely done... I have always strongly believed that serial memory architectures are superior to parallel memory architectures in most every way but my rants have fallen on deaf ears for years....I get the most 'kick out of people knocking the 'latency' problems inherant in serial devices.... (sarcasm) Ask PS2 owners about all of there problems with rambus latency w/their games....I certainly am constantly 'hampered' in games with my i850e/16bit rambus setup (/end sarcasm)
(/end sarcasm)Actually, what happened was that Rambus had filed TOO MUCH info in their original patent filing, which made it too *broad* (not to mention, hard to follow).
The US Patent office asked that they break the original filing into several other patent filings. However it was the US Patent office that wanted this, not Rambus. Rambus didn't care if it was in one filing or twenty. For the purposes of the Rambus v. Infineon, the only patent that mattered was the one defining the bus architecture, which was part of the original 1990 patent filing.
Infineon KNEW this, and this is why Infineon's lawyers HAD to suppress the markman ruling at trial. It was clear that Rambus owned the basic intellectual property for SDRAM, and the only way infineon could win was by suppressing their definition of a bus and removing the only defense to Rambus' own property.
Actually, what I *HOPE* happens here is that people who dislike Rambus SEEK THIS INFO THEMSELVES. Rambus isn't the enemy. They're the only future for memory technology if you want it to scale the way it should.
I don't know why everyone has always hated Rambus. The only problem I have had with their products is that I didn't feel the performane boost justified the cost. However, at this point in time, the difference is becoming a little more noticeable. Rambus is the wave of the future. They are good at what they do. It sounds like JEDEC just wanted to get away with stealing technology from Rambus...
yes let's all praise rambus, who have revolutionized the market with their frequent product releases.
Furthermore, they are clearly the future since all of one companies has seen fit to utilize their product. What were nvidia and ati thinking? don't they know that rdram will improve their performance?
I hate rambus because they were going to increase my bottom line by charging a tax on ram. I've given them enough time to be the wave of the future; if they were going to make something spectactular they would have done it already.
Furthermore, they are clearly the future since all of one companies has seen fit to utilize their product. What were nvidia and ati thinking? don't they know that rdram will improve their performance?
I hate rambus because they were going to increase my bottom line by charging a tax on ram. I've given them enough time to be the wave of the future; if they were going to make something spectactular they would have done it already.
Why spin your engineering wheels releasing product after product when after 2 years of constant releases of DDR-based products can't beat an aging technology?
How does "frequent product releases" fit in with "innovation"? Let's see. With Rambus we have:
RDRAM (PC600 through PC1200 this year)
QRSL (quad data rate)
SerDes
RaSer Cells
And now, their newest emerging technologies:
Yellowstone (100GB/sec, Octal data rates)
Redwood (Far faster than Hypertransport, yet backwards compatible, chip to chip transport)
PCI Express (Bringing it to market with PLX, intel endorsed.)
Ok, let's see what JEDEC has...
DDR. PC1600, PC2100, PC2700 (relying on everyone else to do the rest of the work. They just provide the ram, thankyouverymuch!)
DDR2 for mass market in 2005, and today it can't beat DDR1.
Intel saw it fit with the P4... but then backed off thanks to *GASP* Pressure from JEDEC. I mean, hey... If the big boys in the memory industry said "Hey, look Intel, we're not going to manufacture that RDRAM stuff you want because we don't want to pay the royalty", what do you do? Sacrifice sales? Intel isn't stupid.
NVIDIA and ATI *SHOULD* be thinking Rambus. They have more reason to than anyone...
But oh wait... You have that little problem with JEDEC companies trying to force Rambus out of the market again. Darn.
Why would you want that to happen, btw?
So, what's that tax? 3% of declared manufacturing cost. That amounts to pennies per dram. Not enough to influence prices more than $1 or $2.
But yknow what? I'll tell you what's gonna hurt your bottom line.
The fact that DDR isn't profitable. All these companies are DUMPING DDR at huge losses just to keep it popular. They know full well that NO ONE would want DDR if it costs the same as RDRAM, just like NO ONE would buy a 3000+ Barton if it were priced the same price as the 3.06HT P4. The ONLY thing Team DDR can do is keep price dumping, and hope that Rambus goes away somehow.
Luckily that's not happening according to plan
Team DDR doesn't seem to understand that Rambus is a profitable company because they stay small and lean. Yknow what, you wanna root for a technology, root for one that KEEPS PEOPLE EMPLOYED instead of needing to LAY OFF HALF THEIR WORKERS.
Also, making something spectacular is one thing. Having companies adopt it is another. Rambus is a tiny company. They can't hold an industry hostage like the huge JEDEC megacorporations can.
But alas, JEDEC's time is running short.... tick tick tick...
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 384
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
Facebook
Twitter