RDRAM/RAMBUS - DEAD! Officially. . . . Confirmed?!

[dexvx]

I think rambus will still be in the PC Arena, just at the high end where it offers the most performance. Although Quad channel RDRam is quite interesting (and they say it will cost no more to implement than current dual channel solutions).

I certainly dont see it main-stream or server due to the fact that main-stream likes cheapo crap and server memory you need large quantities of.

 

[Curley]

I agree, it is hard to believe that the best performing memory will no longer be supported by anyone.

Time will tell.

Compute!!!

 

[mchammer187]

Originally posted by: Curley
I agree, it is hard to believe that the best performing memory will no longer be supported by anyone.

Time will tell.

Compute!!!

 

[Deeko]

Wow, I'm suprised...although not particularly upset.

 

[Wolfsraider]

Originally posted by: apoppin

Originally posted by: Wolfsraider
one question then

why did intel just say:

Has DDR overtaken Rambus as the memory of choice for Pentium 4 owners?

We support three memory standards now. We still are firmly behind RAMBUS. It is for the high end, the leadership platform. We appreciate that pricing issues, with RAMBUS memory prices not coming down as much as customers and we would have hoped lead to demand for other RAM variants. So we had to consider people who wanted a price point over performance, we do that through our 845 chipsets.

If you look at the sales figures of our motherboards it clearly shows that DDR based solutions are the most popular. SDRAM is a small percentage, catering for the value market with RAMBUS also being a small percentage catering for the high end, leadership platforms.

here?it doesn't seem to make sense but then few things do.

Ah, you mean the "most HONEST Intel Interview Ever" thread.

Yep, Intel is firmly behind Rambus - one foot is giving them the boot.

rotflmao

just curious as to which way its really going to go
:Q

also i too believe it will be phased out of the destop market but i am hoping that when i get around to building that thar duallie she'll still be an option

 

[sandorski]

Apoppin: Ah see, you're doing this all wrong. When you put "confirmed" in the thread title, it is proper that your first post contain nothing but a claim supporting the title. In fact, no time during the thread are you to provide a link to anything remotely related to the title. Not only did you do it wrong, but your first post was nothing but a link? You're slipping man!


If this turns out to be fact, then all my vitriol towards RDDram will not have been in vain! However, I'll wait a bit before breaking out the champagne.

 

[Wingznut]

Originally posted by: billyjak
Hmm who's right Read

Let's see... Who would I trust more? EEtimes or xbitlabs???

Hmmmm... That's a tough one!

 

[andreasl]

Sure, Intel will eventually validate PC1066 for the 850E chipset, but they have NO other Rambus chipsets on their roadmaps after the 850E. If you want a motherboard that supports PC1333 (the next step after PC1066, due out sometime next year), Serial ATA, AGP 8x etc you'll have to get a SiS chipset instead.

 

[fkloster]

andre...I thought PC-1200 was the next step....have never even heard of PC-1333 :Q

 

[andreasl]

I read somewhere that the next rev of the SiS RDRAM chipset would support it sometime next year. It makes sense with the upcoming 667MHz FSB for Prescott. Only dual channel DDR333 or RDRAM PC1333 could keep it fed.

But it might have been Xbit so take it with a grain of salt...

 

[fkloster]

PC-1333!!! :Q

 

[dexvx]

Let's see... Who would I trust more? EEtimes or xbitlabs???

lol. I think xbitlabs is AMD biased. They did have the lowest p4-2.8 Ghz benchmarks on the net. But anyways, the i850 is still on the Intel roadmap all the way up to 2005. Thats how good the chipset is.

 

[majewski9]

RDRAM was never that great anyways and I am glad to see it go the way of the dinosaur.

 

[andreasl]

Some links so you don't think I'm talking out of my nightcap

Rambus Developer Forum - PC1333

x-bit on upcoming SiS chipsets - SiS R658DX with PC1333 support

VR-Zone confirms, complete with pic and all

 

[apoppin]

Originally posted by: sandorski
Apoppin: Ah see, you're doing this all wrong. When you put "confirmed" in the thread title, it is proper that your first post contain nothing but a claim supporting the title. In fact, no time during the thread are you to provide a link to anything remotely related to the title. Not only did you do it wrong, but your first post was nothing but a link? You're slipping man.
If this turns out to be fact, then all my vitriol towards RDDram will not have been in vain! However, I'll wait a bit before breaking out the champagne

You're thinking of someone else.

Yep, ***CONFIRMED***



The only reason Rambus "exists" is because Intel "pushed" their technology for P4. Now that they have no more support . . . I think you can break out the bubbley.

 

[buleyb]

Originally posted by: majewski9
RDRAM was never that great anyways and I am glad to see it go the way of the dinosaur.

What are you talking about? Sure, Rambus is as shady a company as they come, but the technology is great. Has nobody else realized the shift away from parallel circuits? Serial ATA, PCI-Express, USB (granted serial ports have always been around, but USB is replacing formally parallel interconnects)...RDRAM is just another wonderful serialized technology. This would show more in cost IF motherboard manufactures made more 4layer mobos, AND if more memory manufactures made RDRAM chips (by fault of Rambus and royalties, Micron isn't going to help out anytime soon...too bad).

Rambus may not gain too much share right now, but I highly doubt Intel will let it go, especially considering the performance leads all the benchmarks show with P4 & PC1066.

Rambus will survive as a company (because they do research and patents, not production), and the technology behind RDRAM won't just go away in a magical puff of DDR smoke. DDR is going to get harder and harder for mobo manufactures to get proper timing between banks and to lay the traces. Not to mention the long time before DDR II is expected, and topping out at 533, DDR III better be along soon after...

 

[MercenaryForHire]

Originally posted by: MIGhunter
Not to mention the fact that most of the hogh scores on 3dmark and pcmark are from people using RAMBUS. Of corse there are other considerations but I just don't see that many high scores with DDR.

Yes, because we foolish DDR users sit down to do work and play games, instead of kicking back to unwind with a nice long game of 3DMark2001. Damn, that game is fun. So interactive and indicative of real-world performance, as well as totally unslanted. </sarcasm>

- M4H

 

[AluminumStudios]

As much as I would like to throw in the first shovel of dirt, Rambus is far from dead. Their technology is used in other systems. If I"m not mistaken the PS2 uses Rambus. Also I believe the Cray X1 (formerly codenamed SV2) supercomputer uses Rambus ... and they use a LOT of sticks of it!!!!

Rambus technology isn't just used in PC desktops. This will hurt them, but not kill them.

If it starts looking too bad I'm sure they'll find some other technology to claim as theirs and try to sue everyone again.

This is another case of how poor business practices are killing good technology though ... I'm glad there is another decent contender though (DDR.)

Speaking of DDR I passed Afronova last night ... oh, wait, that's Dance Dance Revoltuion ^_^;

 

[ElFenix]

Originally posted by: Wingznut PEZ

Originally posted by: billyjak
Hmm who's right Read

Let's see... Who would I trust more? EEtimes or xbitlabs???

Hmmmm... That's a tough one!

not dead, but no longer the performance leader, assuming intel doesn't intentionally cripple the chipset somehow.

and actually, that EEtimes article says nothing to contradict the xbit article.

 

[Bobbyeyes]

actually rambus just signed a major licensing agreement w/ intel.
the stock was touted by tec sector analysts(yea i know the tec sector is getting hammered right now).

a paste from the article
i know its is not a memory deal, but rambus will be around for a while

Last month, Rambus RMBS announced they had licensed their RaSer, serdes cell technology to Intel INTC for Ethernet applications. While the details of the arrangement were not disclosed, the deal does not fall under the cross-licensing umbrella agreement Intel and Rambus established in late 2001. Since Rambus is providing the development of the chip for Intel, they will receive additional income for the product.

In addition to the incremental revenue, the deal gives Rambus another nice notch of credibility for RaSer in their belt, and should help in attracting more clients. The potential market for RaSer, as Mr. Hager has mentioned, is significant. The estimates for the numbers of serdes links by 2005 is close to a billion links. Rambus' integrated solution appears to have enough cost saving attributes, and certainly the performance to make it a strong contender in the space, especially since cost-cutting is an order of the foreseeable future for many companies

 

[Valinos]

RDRAM is simply not worth the 10% performance gain over DDR when you knowingly contribute to a company like Rambus that litigates and tries to monopolize everything in their path. They are evil and have horrible anti-competitive business practices. Rambus sucks, therefore RDRAM sucks. It is sad to see such a great technology go down the tubes because of the crap company behind it, but maybe if they didn't have such a bad rep for being dicks and suing everyone then maybe they'd be #1 today. It is their own damn fault and I'm glad to see them go the way of the dinosaur. Hopefully Sony will drop them for their PS3.

 

[Curley]

Originally posted by: Valinos
RDRAM is simply not worth the 10% performance gain over DDR when you knowingly contribute to a company like Rambus that litigates and tries to monopolize everything in their path. They are evil and have horrible anti-competitive business practices. Rambus sucks, therefore RDRAM sucks. It is sad to see such a great technology go down the tubes because of the crap company behind it, but maybe if they didn't have such a bad rep for being dicks and suing everyone then maybe they'd be #1 today. It is their own damn fault and I'm glad to see them go the way of the dinosaur. Hopefully Sony will drop them for their PS3.

Are you talking about Rambus or Microsoft?

 

[Sunny129]

Originally posted by: apoppin
So it looks like Granite Bay - dual channel DDR SDRAM - will be offered as the new performance solution by Intel.

Evidently their agreement with Rambus won't be renewed.

So I can stop considering RDRAM as an alternative "performance" solution. It has no future.

i thought Brookdale was going to be the first chipset to feature dual channel DDR SDRAM. either way, both Brookdale and Granite Bay are dual channel DDR solutions, correct? when are these chipsets going to debut? are there any articles out at hardware webistes that have a review looking at the features of these chipsets and performance of a reference board, or will it be a while before we see anything like that?


as for Rambus, i always chose DDR memory over RDRAM simply due to the cost of memory and also the cost of the mobo needed to run that kind of memory. RDRAM was too expensive, as well as most of the boards needed to run it. but i'm certainly not ecstatic about Intel's phasing RDRAM out of their road map. while DDR memory was by far a cheaper solution, i don't think anyone here will argue that RDRAM was more technologically advanced, and still growing under R&D. all they had to do was come up with a way to make it cheaper to the end user, whether it be cheaper materials, a cheaper manufacturing process, or cheaper marketing schemes. at any rate, it would be nice to see Rambus do something economically viable with their product without the support of the industry giant Intel. without RDRAM competition, DDR industry leaders will be free to do what they want with the price of DDR memory, being the best memory technology out there (if RDRAM truly disappears) with no real competitive technology out there. its like Intel and AMD...while most people favor one over the other, i favor neither. the existance of both companies prevents monopolies for the time being. on top of that, without one or the other, the existing company would not be under pressure to make a better product than its nonoexistant competitor.

 

[apoppin]

Originally posted by: Sunny129

Originally posted by: apoppin
So it looks like Granite Bay - dual channel DDR SDRAM - will be offered as the new performance solution by Intel.

Evidently their agreement with Rambus won't be renewed.

So I can stop considering RDRAM as an alternative "performance" solution. It has no future.

i thought Brookdale was going to be the first chipset to feature dual channel DDR SDRAM. either way, both Brookdale and Granite Bay are dual channel DDR solutions, correct? when are these chipsets going to debut? are there any articles out at hardware webistes that have a review looking at the features of these chipsets and performance of a reference board, or will it be a while before we see anything like that?

You made me look it up.



Granite Bay at IDF!

Although Granite Bay is proposed for high-performance workstations, it seems that the Santa Clara based semiconductor manufacturer is quite liberal about the product positioning, consequently mainboard makers will offer their Granite Bay powered solutions for 1 CPU configurations. Due to the fact that Granite Bay is the first 2-channel DDR SDRAM core-logic from Intel, the industry is waiting for this one impatiently.

Granite Bay supports up to 4GB of PC1600/PC2100 DDR SDRAM, AGP 8x interface and Pentium 4 processors with 400/533MHz FSB and the Hyper-Threading technology enabled. Intel?s Granite Bay will feature ICH4 South Bridge, hence, we need to add USB 2.0, ATA-100 and 32-bit 33MHz PCI support as well.

 

[MIGhunter]

Originally posted by: MercenaryForHire

Originally posted by: MIGhunter
Not to mention the fact that most of the hogh scores on 3dmark and pcmark are from people using RAMBUS. Of corse there are other considerations but I just don't see that many high scores with DDR.

Yes, because we foolish DDR users sit down to do work and play games, instead of kicking back to unwind with a nice long game of 3DMark2001. Damn, that game is fun. So interactive and indicative of real-world performance, as well as totally unslanted. </sarcasm>

- M4H

Yeah, and I drive my car too. That doesn't mean that I didn't look at the MPG and saftey ratings befor I bought it. Not to mention it is a good measurement of functionality. If video games gave you a score I would post them. Besides I wanted something fast for chemistry applications.