Rambus is up to their old tricks again
- Thread starter Bateluer
- Start date
<<Maybe they should concentrate on improving their shoddy technology>>
Another noob with no knowledge of RDRAM. Apparently you also missed this quote from the very article you supplied
<<on its own RDRAM technology, which most people in the industry believe is technologically superior.>>
Yeah it's shoddy. Thats the ticket
Another noob with no knowledge of RDRAM. Apparently you also missed this quote from the very article you supplied
<<on its own RDRAM technology, which most people in the industry believe is technologically superior.>>
Yeah it's shoddy. Thats the ticket
People wouldn't hate Rambus if Rambus wouldn't be so shady AS A COMPANY. They pulled a lot of crap that pissed people off with the whole JEDEC debacle.
Probably some of their execs use to work for Enron.
Probably some of their execs use to work for Enron.
<< Memory type is good only if its bandwidth is good. RDRAM is all about bandwidth. It means that RDRAM is good. >>
Not necessarily. Latency matters largely with Athlon's, and even with P4's it matters. I do agree that RDRAM is good stuff for the P4, but RDRAM isn't for every CPU,
<< This certainly beyond a shadow of a doubt is a very uninformed and non-intelligent thread...
>>Don't get me started fkloster... :|
Hehe, just kidding.
Yeah, RDRAM as a technology is pretty decent, it would be far better if it had a larger bus (16bits bus is 1/4 of what SDRAM/DDRRAM has) and the latency is poor, but it's obscenely fast, so that makes up for it's shortcomings. If you look at RDRAM vs. DDR with comparable bandwidths, there is a slight edge to RDRAM - but that could also be due to motherboards. RDRAM is also a bit more expensive - the price gap used to be much larger, but it has recently shrunk.
However, the main reason to hate RDRAM is because of the company that owns the technology - Rambus. A company with shoddy business ethics that will do anything (in court) to get ahead. That could include stealing open technology, patenting said technology, breaking contract with JEDEC, claiming royalties on said technology, etc.
Sickening the kind of business that they run.
cmdrdredd
Lifer
- Dec 12, 2001
- 27,052
- 357
- 126
I always hated RDRAM because it seemed to be like SuperCD...a standard that never was.
Not to mention that if RAMBUS was so good why would Nvidia not use it?
BTW: I think RDRAM's major flaw is that it has to be installed in Pairs...eek, and isn't there supposed to be DDR400 soon? I read something that Nvidia had a DDR400 motherboard on display at some show somewhere...I really can't remember where I read it though. Anyway...DDR400 will probably woop RDRAM. Why do I say that? Well...because it's pretty close now.
Not to mention that if RAMBUS was so good why would Nvidia not use it?
BTW: I think RDRAM's major flaw is that it has to be installed in Pairs...eek, and isn't there supposed to be DDR400 soon? I read something that Nvidia had a DDR400 motherboard on display at some show somewhere...I really can't remember where I read it though. Anyway...DDR400 will probably woop RDRAM. Why do I say that? Well...because it's pretty close now.
<< However, the main reason to hate RDRAM is because of the company that owns the technology - Rambus. >>
<< BTW: I think RDRAM's major flaw is that it has to be installed in Pairs...eek, and isn't there supposed to be DDR400 soon? I read something that Nvidia had a DDR400 motherboard on display at some show somewhere...I really can't remember where I read it though. Anyway...DDR400 will probably woop RDRAM. Why do I say that? Well...because it's pretty close now. >>
Needing to be installed in pairs isn't an issue of RDRAM, it has to do with the fact that 850 uses a Dual Channel Memory controller. It's not limited to RDRAM. nForce, a Dual Channel DDR chipset, requires the memory to be installed in pairs.
<< it would be far better if it had a larger bus (16bits bus is 1/4 of what SDRAM/DDRRAM has) and the latency is poor, >>
That does not make a whole lot of sense. The reason RDRAM achieves it's bandwidth is because of the 16-bit bus. I'd rather it stay that way. As for the latency, that is bull. PC1066 RDRAM Latency on 850 is on par with PC2100 on 845-D.
Wolfsraider
Diamond Member
- Jan 27, 2002
- 8,305
- 0
- 76
<<
<< BTW: I think RDRAM's major flaw is that it has to be installed in Pairs...eek, and isn't there supposed to be DDR400 soon? I read something that Nvidia had a DDR400 motherboard on display at some show somewhere...I really can't remember where I read it though. Anyway...DDR400 will probably woop RDRAM. Why do I say that? Well...because it's pretty close now. >>
Needing to be installed in pairs isn't an issue of RDRAM, it has to do with the fact that 850 uses a Dual Channel Memory controller. It's not limited to RDRAM. nForce, a Dual Channel DDR chipset, requires the memory to be installed in pairs.
<< it would be far better if it had a larger bus (16bits bus is 1/4 of what SDRAM/DDRRAM has) and the latency is poor, >>
That does not make a whole lot of sense. The reason RDRAM achieves it's bandwidth is because of the 16-bit bus. I'd rather it stay that way. As for the latency, that is bull. PC1066 RDRAM Latency on 850 is on par with PC2100 on 845-D. >>
and um isnt ddr2 also t be installed in pairs?
dullard
Elite Member
- May 21, 2001
- 25,889
- 4,471
- 126
"I think RDRAM's major flaw is that it has to be installed in Pairs."
Why would anyone care if it is installed in singles, pairs, or 10's? All that you have to do to overcome it is to have motherboard manufacturers place more slots on the MB. RDRAM motherboards typically have 4 slots - and capability for 8 RIMMS is common in workstations. DDR motherboards generally have 2 or 3. So since there are more slots, why care how many need to be installed? The extra slots add a few dollars to the motherboard cost, but that is insignificant when compared to the rest of the computer price.
"Latency matters largely with Athlon's, and even with P4's it matters"
You know full well that the PC1066 that will be released in the next week or two has a drastic improvement in latency - lower than most DDR and equal to the best DDR.
"Not to mention that if RAMBUS was so good why would Nvidia not use it?"
RDRAM has been put in some video cards. However at the time, RDRAM was so expensive that it wasn't worth the money you had to pay. Even with price drops RDRAM has never been cheaper than DDR - so there hasn't been a monetary reason to develop RDRAM video cards. RDRAM has only reached reasonable prices in the last 6 months - there hasn't been enough time for a new video card to be developed with this knowledge in mind.
"Maybe they should concentrate on improving their shoddy technology"
Isn't the PC1066 an improvement? What about the PC1200 demoed a few months ago - ready to be used as soon as a processor needs that bandwidth? What about 32-bit RDRAM that won't need to be used in pairs even with dual channel memory controllers (Intel dropped this due to the reasons I stated above)? They have been making improvements. By the way have you read the patent? If so what in particular do you think is incorrect in the patent? If not, why are you complaining about something that you haven't read?
Why would anyone care if it is installed in singles, pairs, or 10's? All that you have to do to overcome it is to have motherboard manufacturers place more slots on the MB. RDRAM motherboards typically have 4 slots - and capability for 8 RIMMS is common in workstations. DDR motherboards generally have 2 or 3. So since there are more slots, why care how many need to be installed? The extra slots add a few dollars to the motherboard cost, but that is insignificant when compared to the rest of the computer price.
"Latency matters largely with Athlon's, and even with P4's it matters"
You know full well that the PC1066 that will be released in the next week or two has a drastic improvement in latency - lower than most DDR and equal to the best DDR.
"Not to mention that if RAMBUS was so good why would Nvidia not use it?"
RDRAM has been put in some video cards. However at the time, RDRAM was so expensive that it wasn't worth the money you had to pay. Even with price drops RDRAM has never been cheaper than DDR - so there hasn't been a monetary reason to develop RDRAM video cards. RDRAM has only reached reasonable prices in the last 6 months - there hasn't been enough time for a new video card to be developed with this knowledge in mind.
"Maybe they should concentrate on improving their shoddy technology"
Isn't the PC1066 an improvement? What about the PC1200 demoed a few months ago - ready to be used as soon as a processor needs that bandwidth? What about 32-bit RDRAM that won't need to be used in pairs even with dual channel memory controllers (Intel dropped this due to the reasons I stated above)? They have been making improvements. By the way have you read the patent? If so what in particular do you think is incorrect in the patent? If not, why are you complaining about something that you haven't read?
BD231
Lifer
- Feb 26, 2001
- 10,568
- 138
- 106
"Latency matters largely with Athlon's, and even with P4's it matters"
You know full well that the PC1066 that will be released in the next week or two has a drastic improvement in latency - lower than most DDR and equal to the best DDR.
Last I heard latency was becoming more of an issue with DDR than RDRam. The higher you clock DDR the higher the latency get's correct?
You know full well that the PC1066 that will be released in the next week or two has a drastic improvement in latency - lower than most DDR and equal to the best DDR.
Last I heard latency was becoming more of an issue with DDR than RDRam. The higher you clock DDR the higher the latency get's correct?
LOL I just love it when people see some computer companies as "good" and others "bad"
They don't CARE about you.
They both just want your MONEY.
They don't CARE about you.
They both just want your MONEY.
LOL I just love it when people see some computer companies as "good" and others "bad"
They don't CARE about you.
They both just want your MONEY.
True. And in return for my money I want a solid component. Everyone out there (at least for the most part) has a choice of whether a certain product is worth what a company says it is.
Only without competition does a company's true merits become clear.
Anyhow, all of said companies have more money than me, so I hate them all. :disgust:
They don't CARE about you.
They both just want your MONEY.
True. And in return for my money I want a solid component. Everyone out there (at least for the most part) has a choice of whether a certain product is worth what a company says it is.
Only without competition does a company's true merits become clear.
Anyhow, all of said companies have more money than me, so I hate them all. :disgust:
BD231
Lifer
- Feb 26, 2001
- 10,568
- 138
- 106
I RDram may be good stuff, but I aint buying it. and yes, I make that decsion off the basis that the company is suck.
So your saying that the screwy ideals of some dried up old men stop you from using it, even though they have an all around competitive product in terms of price AND performance?
So your saying that the screwy ideals of some dried up old men stop you from using it, even though they have an all around competitive product in terms of price AND performance?
<< So your saying that the screwy ideals of some dried up old men stop you from using it, even though they have an all around competitive product in terms of price AND performance? >>
Thats right. What's your point?
If you had a patent on something would you let others profit from it without paying you? If your answer is yes your lying. Rambus has done nothing wrong in suing companies they think have infringed on their patents. Us a AMD patent without paying them and see don't they sue your ass.
dullard
Elite Member
- May 21, 2001
- 25,889
- 4,471
- 126
<< Rambus has done nothing wrong in suing companies they think have infringed on their patents. >>
You are semi-right and semi-wrong. The latest lawsuit results have come to these conclusions (note as more appeals are finalized the results might change):
1) Rambus was legal in obtaining its patents.
2) Rambus had the right to sue for infringement.
3) Rambus neglected to tell the other companies about the patents (which is illegal - so Rambus did do something wrong).
4) Since Rambus neglected to tell the other companies about the patents, they are banned from suing in the US for past infringments on their legal patents (but may sue for future infringements).
5) Rambus may legally sue outside the US for those infringements.
If Rambus didn't do step (3) then you would be right that Rambus did nothing wrong.
<< Rambus neglected to tell the other companies about the patents... >>
Am I incorrect in my assumption that patents must be 'filed' and documented? Otherwise they are not patents. If they were filed and correctly patented, then Rambus would have 0% obligation to notify other memory manufacturers of the infringments. Patent review would fall squarly on the shoulders of those who whould be desinging DDR without consulting existing memory architecture patents...
dullard
Elite Member
- May 21, 2001
- 25,889
- 4,471
- 126
<< If they were filed and correctly patented, then Rambus would have 0% obligation to notify other memory manufacturers of the infringments. >>
If you have a patent, and are asked if you have a patent then legally you must answer yes. Rambus answered 'no' - causing other companies to invest time and money in this technology thinking it was an open standard. Rambus had the patent the whole time and did nothing to stop them from infringing on it. That is where Rambus went wrong (legally).
I do see your point that the other companies could have just did a patent search and found out that it was patented - and the whole problem would have never existed.
Latency is a big matter on RAM. Because rambus is in serial, the more ram you have the overall latency of it increases. Thats why its NOT good to be used in server applications that demand a huge amount of ram.
However, Rambus latency can actually be *lower* than DDR SDram latency. Once your throughput reaches anywhere near max, the latency of SDram increases. It was a long time ago when Intel did a comparison to i820 and i440BX, when memory bandwidth consumption was relatively low, the i440BX outperfromed the i820 by a large margin, latency wise. However, once it got near the half way mark, it was about even, and anywhere after that the i820 outperformed the i440BX latency wise. When memory bandwidth was near max capacity, the RDram would significantly outperform the i440BX's SDRam platform. The same situation would apply to DDR SDram because it is a derivation of original SDram. The very ironic thing was that i820 was a VERY poor implementation of RDram (only single channel) and i440BX was probably the best implementation of SDram.
As for the nForce hype, its largely hype. I mean when people were talking about nForce, they made it seem that Dual channel DDR will bring godly performance to the Athlon platform. Now look at it, about a year later, dual channel DDR on nForce barely beats single channel DDR on via KT266a platform in performance AND memory bandwidth.
Btw, RDram should be called DDR RDram because it is a double data rate. When people say 'DDR' they synonomously think its DDR SDram, but the terminology can be applied to RDram because it is technically, DDR RDram.
I think that Rambus itself has some merits in its lawsuits. They were the FIRST to develop DDR technology and to utilize it on their RDram. DDR SDram came afterwards. Beyond that, I wouldnt know.
However, Rambus latency can actually be *lower* than DDR SDram latency. Once your throughput reaches anywhere near max, the latency of SDram increases. It was a long time ago when Intel did a comparison to i820 and i440BX, when memory bandwidth consumption was relatively low, the i440BX outperfromed the i820 by a large margin, latency wise. However, once it got near the half way mark, it was about even, and anywhere after that the i820 outperformed the i440BX latency wise. When memory bandwidth was near max capacity, the RDram would significantly outperform the i440BX's SDRam platform. The same situation would apply to DDR SDram because it is a derivation of original SDram. The very ironic thing was that i820 was a VERY poor implementation of RDram (only single channel) and i440BX was probably the best implementation of SDram.
As for the nForce hype, its largely hype. I mean when people were talking about nForce, they made it seem that Dual channel DDR will bring godly performance to the Athlon platform. Now look at it, about a year later, dual channel DDR on nForce barely beats single channel DDR on via KT266a platform in performance AND memory bandwidth.
Btw, RDram should be called DDR RDram because it is a double data rate. When people say 'DDR' they synonomously think its DDR SDram, but the terminology can be applied to RDram because it is technically, DDR RDram.
I think that Rambus itself has some merits in its lawsuits. They were the FIRST to develop DDR technology and to utilize it on their RDram. DDR SDram came afterwards. Beyond that, I wouldnt know.
-----
Latency is a big matter on RAM. Because rambus is in serial, the more ram you have the overall latency of it increases. Thats why its NOT good to be used in server applications that demand a huge amount of ram.
However, Rambus latency can actually be *lower* than DDR SDram latency. Once your throughput reaches anywhere near max, the latency of SDram increases. It was a long time ago when Intel did a comparison to i820 and i440BX, when memory bandwidth consumption was relatively low, the i440BX outperfromed the i820 by a large margin, latency wise. However, once it got near the half way mark, it was about even, and anywhere after that the i820 outperformed the i440BX latency wise. When memory bandwidth was near max capacity, the RDram would significantly outperform the i440BX's SDRam platform. The same situation would apply to DDR SDram because it is a derivation of original SDram. The very ironic thing was that i820 was a VERY poor implementation of RDram (only single channel) and i440BX was probably the best implementation of SDram.
-----
This Fud is still floating around, my god.
come on people, latecny does not magically increase because you ask your ram for more data/second. This is completly false and anyone who beleives otherwise has fallen victum to some very shady PR by Intel, rambus and possibly Anand ( though i suspect Anand just does not know any better and is not really in on it )
The information your getting this from is based of something a little dishonest on Intels part. What They did was basically ask the ram to deleiver more data/sec that it's theoritcall maxium, resulting in a sustained effciency of infinity ( or 0% depending on how you view it ). What of course they ommit is, RD-RAM will do exactly the same thing provided you ask it to give you more than it's theorietical max. In the graph used, Intel does not get to close to RD-RAM maxium bandwdth potential.
If you star at this Intel PR long enough, you will not only realize Intel is basically really streching the truth, but are even dumb enough to break the laws of mathmatics and the universe to tell the lie. Hint. Intel believes if you reach infinity and keep going, you suddendly start going backwards. Last I heard this is in the realm of a very grey area in theoritical mathmatics/Positive-physics which has no real life application known yet. Excpet at intel of course!
Latency is a big matter on RAM. Because rambus is in serial, the more ram you have the overall latency of it increases. Thats why its NOT good to be used in server applications that demand a huge amount of ram.
However, Rambus latency can actually be *lower* than DDR SDram latency. Once your throughput reaches anywhere near max, the latency of SDram increases. It was a long time ago when Intel did a comparison to i820 and i440BX, when memory bandwidth consumption was relatively low, the i440BX outperfromed the i820 by a large margin, latency wise. However, once it got near the half way mark, it was about even, and anywhere after that the i820 outperformed the i440BX latency wise. When memory bandwidth was near max capacity, the RDram would significantly outperform the i440BX's SDRam platform. The same situation would apply to DDR SDram because it is a derivation of original SDram. The very ironic thing was that i820 was a VERY poor implementation of RDram (only single channel) and i440BX was probably the best implementation of SDram.
-----
This Fud is still floating around, my god.
come on people, latecny does not magically increase because you ask your ram for more data/second. This is completly false and anyone who beleives otherwise has fallen victum to some very shady PR by Intel, rambus and possibly Anand ( though i suspect Anand just does not know any better and is not really in on it )
The information your getting this from is based of something a little dishonest on Intels part. What They did was basically ask the ram to deleiver more data/sec that it's theoritcall maxium, resulting in a sustained effciency of infinity ( or 0% depending on how you view it ). What of course they ommit is, RD-RAM will do exactly the same thing provided you ask it to give you more than it's theorietical max. In the graph used, Intel does not get to close to RD-RAM maxium bandwdth potential.
If you star at this Intel PR long enough, you will not only realize Intel is basically really streching the truth, but are even dumb enough to break the laws of mathmatics and the universe to tell the lie. Hint. Intel believes if you reach infinity and keep going, you suddendly start going backwards. Last I heard this is in the realm of a very grey area in theoritical mathmatics/Positive-physics which has no real life application known yet. Excpet at intel of course!
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter