Slashdot Mirror
Intel To Rambus: Long Walk, Short Pier
NightHwk writes: "A article here reports: "We made a big bet on Rambus and it did not work out," admitted Craig Barrett, Intel chief executive. "In retrospect, it was a mistake to be dependent on a third party for a technology that gates your performance." [...] "We hoped we were partners with a company that would concentrate on technology innovation rather than seeking to collect a toll from other companies," Barrett said.
"
Yes.
.
Very few people have been as vocal a critic of intel as I have. But after this statement, it shows an attitude of almost . . . human, humility.
Have our harsh anti-intel sentiments published on line finally gotten through to someone who makes decisions at intel? Are there really decent human beings in positions of power in corporate America?
I could only hope that Motorola can find enough grease for it's neck so they can pull their heads out of their asses as well. Otherwise, I might actually *sign* that OSXonIntel.com petition. .
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Either that, or he figured out that his posting was going to be sid#25, and put in the link manually
Bad slashbot. Bad. Sit! Giggle!
Read the rest of this comment...
`ø,,ø`ø,,ø!
Free Software: Like love, it grows best when given away.
I could have sworn what he really said was...
"It was a mistake to rely on Gates for performance"
Hmmm....
. . . and I guess Apple will now license RAMBUS technology for Macintoshes.
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Rambus multiplexes the heck out of everything, to get lower pin counts. Address selection involves sending a string of values accross the bus. Data comes back in a burst of values. To get even similar performance, the Rambus architecture has to run their narrower bus several times faster and faster parallel busses have more problems - their harder to synchronize, have higher energy emissions and hence potentially more cross-talk and they are subject to resonant interference at shorter distances. This last was the problem that tubed the Intel three-slot controllers. Rambus thought they were clever enough to work around these problems, which aparantly they weren't.
They also seem to have failed to find out the frequency of different memory access modes (burst sequntial vs. random, what lengths predominate etc.) in a typical system, so they produced a system tuned for long sequential data burst. This resulted in high latencies which kill Rambus performance in typical useage (though it should really fly in streaming applications).
Rambus has a long list of nasties. Any one of them might be annoying, but collectively they make the whole thing a house of cards. Issues:
There are other problems, but that's a good starting point.
Lacking <sarcasm> tags,
RAMBUS has gone for high frequencies: 800 MHz x 16bits = 1.6 GB/s peak. But 800 MHz isn't easy to design onto working PCBs. Tough enough on silicon dies, and mobo's had plenty of grief just getting to 100 MHz a few years ago. 800 is a big stepout.
Every trace has to be fully simulated, none of the usual cut'n'try. Can you say data-dependant crosstalk? Sure you can! Getting close to where you need to use wave-guide techniques.
The current slew rate (Amps/ns) to drive the traces is also proportionately higher, and it isn't easy to build such low ESR capacitance.
Just because you can run 100 MHz over 100m of clothesline (oops, I mean quality Cat5 cable) doesn't mean you can run 800 MHz over 10cm of traces. Balanced lines might help, but then you double the pincount.
Intel's contract with Rambus forces them to ship motherboards supporting Rambus DRAM... but it doesn't say HOW it has to support them.
Intel's chipset plans call for the inclusion of a separate memory bank that will make it possible to increase the performance of the chipset-embedded graphics core by adding graphics-only RAM to this special bank. The special bank will be Rambus, the main memory will be double data rate RAM.
The kicker is that NO ONE will EVER use the special slot. The embedded graphics core is dog slow, and only marginally less so with the dedicated RAM added.
Any one wanting more performance will mearly bypass the embedded core instead of actually using Rambus, but Intel will have shipped a motherboard with a Rambus interface, fulfilling their contractual obligation.
Tastes Like Chicken
Only embedded applications (like a Nintendo 64) use a direct interface with the RAM. It is not a good idea to create the memory interface directly on the CPU if you want third parties to be able to make motherboards (and chipsets too). You would have to have a chip for RDRAM and a chip for SDRAM, and possibly a chip for DDR-SDRAM, and they would not be interchangeable. Intel used the right implementation, but for the wrong technology.
However, there is something of a silver lining for Intel. The company's investments in Micron and Samsung have produced a generous return as the shares have appreciated significantly in value.
I suspect this investment gain is not as rosey as this author is claiming it is.
That which does not kill me only makes me whinier
Rambus is able to run the bus much faster because they use a serial protocol, along with an interesting clocking scheme. They have lower pin counts than a parellel bus with equivalent bandwidth, but then you can have multiple rambus channels for an equivalent pin count and much higher bandwidth. Yes, there are signifigant pcb problems for such a tight bus, which is why the PC world isi probably not ready for rambus yet.
There's an analysis of this announcement on The Register, here
Short Summary: Intel might be regretting it now, but they're legally committed anyway...
--
I'd rather have a bottle in front of me than a frontal lobotomy
Bill - aka taniwha
--
Bill - aka taniwha
--
Leave others their otherness. -- Aratak
In the immortal words of CaptKos, "You pay for the drive, or you pay for the media. This is always true.".
He was referring to tape drives at the time; I was commenting that DAT drives were expensive, but the tapes were much cheaper than their Travan counterparts. However, this really seems to be a truism in computing.
This is unlike cars, where high-end models typically require more expensive high octane fuel. Instead, think of laser printers that cost an arm and a leg but use dirt cheap toner, or the case of winmodems where you trade a cheap DSP for cycles from an expensive CPU.
It seems to almost never work out were the cheap solution is less expensive in the long run, except for SCSI vs. IDE and Intel's overpriced CPUs vs. cheap Athlons.
Dewey, what part of this looks like authorities should be involved?
Well - atleast the CEO seems to be able to own up and say 'Yes we screwed up and yes we realise this', rather than the obfuscated search for a scape goat that most other companies seem to participate in these days. Hopefully this will be a trend for the future. Would be nice to hear more people admit blame and move on trying to change it than do everything they can do cover over mistakes.
Can't Intel theoretically release a P4 chip, chipset, motherboard, etc, for RDRAM, and then an otherwise identical solution, but otherwise call it PV with DDR-SDRAM support and maybe a few other goodies, and just flush Rambus down the toilet?
The nick is a joke! Really!
GPL Deconstructed
When was the last time a *large* company came out and said "We made a bone-head manuever. We really wish our partner would pull it's head out of it's anal passageways and stop trying to screw everybody else."??? Maybe I'm having short-term memory problems, but I can't think of a single time that a company has been this fothright about something that *big*. I'm still going to be buying a Duron, mind you (hell, it's on the way), but my opinion of Intel has just risen a couple notches. Kudo's to you!
Well, looks like Rambus did serve as a positive force after all. Because of Rambus, Intel has lost *serious* ground on the low-end processor front. Canceling Timna was a major setback and how long do we expect the Celeron line to last under increased pressure from dirt-cheap Athlons? Because of Intel's dalliance with Rambus, now AMD has a major boost in the low-end of the processor market. Couple that with the fact it isnt easy to overclock newer Celerons anymore and soon every PC under $1000 is gonna be AMD Inside :) Sure, Intel still owns the top end of the server market but its the low end that really matters. Cheap PC's are heading towards telephone status in terms of ubiquity - we aren't there yet by a longshot, but I think we are on that path.
Don't blame me - I voted for Howard Dean. http://dean2004.blogspot.com
The article failed to mention that Rambus is significantly inferior to DDR SDRAM. The industry isn't rejecting Rambus just because of price, its a price per performance ratio. Why pay more for Rambus its 5-15% slower than even pc133. Intel made a serious wrong turn. DDR SDRAM on an Athlon running linux will be the END of "Wintel" =^)
"I know where you wanted to go today, But we decided to stop here instead!"
Personally, I'm looking forward to NVidia's chipset if they do actually make it. It would be the first non-VIA, natively-SDRAM chipset for an Intel processor since the 440BX.
"Ancillary does not mean you get to rule the world." --U.S. Circuit Judge Harry Edwards, speaking to the FCC's lawyer
What, Intel can't release P4 with Rambus support, and then a quarter later, release the P5 with DDR-SDRAM support? Or the P4a?
The nick is a joke! Really!
GPL Deconstructed
Intel nailed their own butt to the wall with this one, the P4, so long as there is a Rambus, is exclusively (in so far as Intel brand motherboards and chipsets) RDRAM.
Probably the motive behind Barret's public pissing on Rambus is to discourage Rambus from pursuing these revenue sources which only hurt Intel (as Intel would rather not be doing the RDRAM dance.) Rambus hurts their patron at their own expense, as Intel should be moving millions of memory hungry boxes. Rambus wants to have their cake and eat it, too.
It could be interesting if Intel were to launch a legal salvo at Rambus. Something to the effect of working against Intel, bad faith, that sort of thing. Rambus may be able to withstand the assault from Micron and Infineon, but Intel would probably inflict a mortal wound. Perhaps the grumbling is meant as the warning growl of a bear, procede toward her cubs at your own peril.
--
Chief Frog Inspector
A feeling of having made the same mistake before: Deja Foobar
That's how every company operates. Unfortunately, under the "safe harbor" clause "Under The Private Securities Litigation Reform Act of 1995", they're protected from litigation based on the statements. You can barely even say "N'yah, n'yah! I told you so!" without having their lawyers/liars knocking on your door.
"Ancillary does not mean you get to rule the world." --U.S. Circuit Judge Harry Edwards, speaking to the FCC's lawyer
They have to support RAMBUS in the P4, but do they have to only support RAMBUS? I assume they can still support other memory technologies as well. Anyone know?
Oh yup - you're right there - with the name bit - wouldn't know about their peripherals bing overpriced though.
Matt Thompson - Actuality - Insert product here.
QDR SRAM is not meant as a replacement for main memory. Like most SRAM, it's meant for high speed buffers, such as on a network router, hard drive, or (maybe) processor L2 cache (though it only runs at 200 MHz). SRAM is much more expensive and consumes much more power than DRAM, including DDR SDRAM and RDRAM.
DDD SRAM != DDR SDRAM. Read Ars Technica's RAM guide for a good description of the underlying differences in design between SRAM and SDRAM.
Also, you'll note that QDR SRAM does not violate the most contentious of Rambus Inc.'s patents because it's older SRAM technology, not SDRAM. They aren't really doing anything that new. All QDR SRAM is is double-ported DDR SRAM. All the information is there on the website you linked to.
If it's for-profit but free, you're not the customer -- you're the product (e.g., the Slashdot Beta's "audience").
The busness world has been long notable for their CYA practices. Why buy Dell when you can fabricate your own machines for less than 1/2 the price of the Dell machine? Well, if it screws up, then everybody else does the same thing... Same with Windoze, and Intel. Most folks would rather be in the crowd rather than in the right. Unfortunately for Intel, they are losing market share in Europe (especially Germany) to AMD. You can fool some of the people all of the time and all of the people most of the time, but sooner or later the truth will out.
"Windows is fairly widely used but we all know that it's not necessarily better..."
I'm going to nit-pick.
This statement depends on your definition of "better".
Like you said, it is quite possible that Rambus memory is better in the context of a Playstation 2, but that doesn't mean it is better in the context of a PC.
Similar concept goes along with Windows being better. To the majority of consumers who have bought it, it is better. Otherwise they would be buying competing products and Apple wouldn't be reporting slow sales.
"We apologize for making keyboards without the letter 'Q'. We're sorry. We're morons. We hear strange voices in our heads. I have broccoli in my socks."
Dilbert (to Dogbert): "Good writing."Dogbert: "Thanks."
"Ancillary does not mean you get to rule the world." --U.S. Circuit Judge Harry Edwards, speaking to the FCC's lawyer
Note to Taco, Hemos et al: Please don't link articles from ft.com any more.
--
Build a man a fire, and he's warm for a day.
Time is Nature's way of keeping everything from happening at once... the bitch.
In case anyone thinks that Intel backing away from Rambus is a sign they're becoming nice guys, read the above comment and the other responses to it closely: Intel is most likely backing away from Rambus for technical reasons, not because of the tollkeeper business model. Intel has realized that RDRAM will drive up design and manufacturing costs and thus make it much harder to price products competitively. What would happen to Apple's market share if X86-based PCs suddenly got significantly more expensive?
My guess is that Intel knew about the technical problems with RDRAM early on, but believed that as the technology matured, those problems could be eliminated - manufacturing costs would drop, and hardware designers would learn how to use the technology effectively. This belief turned out to be a mistake.
Barrett's rhetoric about Intel's disgruntlement with Rambus's IP strategy is just a cover story: By focusing on that issue, he hopes to draw attention away from the fact that Intel apparently made a faulty technical analysis when deciding to push RDRAM. A perception that Intel had made a technical mistake would be far more damaging to them than a perception that they had partnered with what turned out to be a pack of jerks.
Complements of a lead from "JC's", here is an interesting article about Intel and DDR. Rambus always impressed me as an idea looking for a problem. Given Intel's switcheroo, it may be a problem looking for a wastebasket.
Rambust reported a 9 cent quarter yesterday. With a street estimate of 47 cents for the year going forward, they are at a price/earnings of 150. With Intel going more and more to DDR, etc. and suits like Hyundai's, I wonder if that 47 cents isn't a bit on the optimistic side...
While I don't root for Intel, its good to see that they finally realized that they are riding a lame horse with this one...
Rambus not signing any more SDRAM licences, combined with estimated legal costs, suggests a couple things
They intend to shut off the license-revene spigot on certain technologies to push others(e.g. RDRAM
They are determined to self destruct through litigation the way Ashton-Tate did.
As much ire is directed at Rambus, I can't see the logic behind this. Ashton-Tate is relevant because they channeled resources into legal wrangling, rather than R&D and customer service. They failed, by putting all their eggs into one basket, a very wrong basket.
Rambus seems to be betting the farm on sales of RDRAM and winning suits against DDRAM manufacturers (Micron, Infineon, et al) By the time appeals have run their course Rambus will probably have borrowed deeply to continue litigation and have spent little on R&D or IP acquisitions. (I wouldn't sell my Patent;¹ to a company which looks lik they may never be able to pay royalties.)
;)
That is, assuming I was some evil scum with a patent
--
Chief Frog Inspector
A feeling of having made the same mistake before: Deja Foobar
I guess this means that Intel will now use DDR SRAM for upcomming chipsets. And then possibly QDR SRAM
*drool*
--
Its nice to see that Intel is realizing its mistakes. Perhaps they aren't quite the bumbling behemoth that we've seen them as for the last year or so. Realizing their mistakes is the first step, now they just have to remedy them.
The article makes a very proper assumption that Intel's partnership with Rambus has been a good part of its downfall. Its not enough that their processors are more expensive than their competitors', but if you pair that up with a system that is required to use Rambus RAM, the price of that system comes up quite a bit higher than its nearest AMD-equipped neighbor. No doubt if they had endorsed a more Open RAM standard such as DDR they would be in much better shape.
But its not only that, Rambus' lawsuit frenzy has shown badly on Intel as well, since they seemed to be the only one supporting the company in the entire industry. This is most likely why those other RAM companys signed up with Rambus to pay them royaltys--because they believed the Behemoth, Intel, wanted Rambus to be on top--and they didn't want to get on Intels bad side.
But the fact that Intels CEO has "badmouthed" Rambus shows that he may be more on top of the situation than we thought...and their may be hope for the future yet. Sure, its nice to see AMD catching up and giving Intel some definite competition, but I don't think any of us want to see them fall completely out of the race.
The only thing I wonder now is for the Pentium 4. The article states that they have an agreement to include "support" for Rambus with the processor...but does that mean that they cannot include support for other, better forms of ram, such as DDR? I suppose that will show how all of this will turn out.
In any case...this is a start.
-Julius X
-Julius X
remove "-whatkindofspamdoyoutakemefor-" from email to send
PC prices are dropping. More importantly is that for your dollar you get much more than you got last year, or the year before, etc. Memory is only one component, which Rambus seems intent on keeping the price high on, by demanding an IP license cut. The drops in CPU prices in the last month easily offset whatever Rambus is squeezing out of you for passed-on-to-the-consumer costs.
Don't, however, assume this will play the same for the P4 when it rolls out (assuming it can stay on track this once) Nov. 20th. Even with Intel cutting the initial prices for P4's and subsidizing PC makers (basically giving them the premium for using RDRAM, makes Intel look all the more foolish, eh?) they'll still be for the deep of pockets.
--
Chief Frog Inspector
A feeling of having made the same mistake before: Deja Foobar
Intel/Rambus strategy: Oh, sure, you can keep using your SDRAM, but then you take a major performance hit.
They both bribe you to give them more money.
Intel HAS to support rambus in the P4. But they're unhappy about it.
Not quite a long walk off a short pier, but probably after the P4 intel will learn to do a little more research before making big decisions.
-- IANAEG - I am not an elder god.
Rambus is two things - it's the name of the company and that name of the technology they designed to, allegedly, speed up PCs.
They do design work but not actual manufacture of chips - this is left to partners such as Kensington and Viking etc.
Recently Rambus have claimed that they hold patents that affect other RAM technologies such as SDRAM which has angered other RAM makers.
Matt Thompson - Actuality - Insert product here.
Hoped! Classic! Biggest chip manufacturer in the world basing its future on dreams...
With planning like that, i bet AMD are pissing themselves!
This is good news! Intel must have finally gotten around to reading Tom's Hardware.
Maybe it's the beginning of them getting their head out of their arse. I wonder how much behind-the-scenes wrangling there has been between Intel ("make your damn memory work") and Rambus ("give us money") has gone one leading to this moment. Maybe Intel Management woke up during one of the meetings and said, "Hey! These guys are a law firm disguised as a hardware company! Argh!"
Anyway, I'm happy about this. Now, if Intel can just help shut down the Rambus crusade against all other memory makers, it'll be perfect. Find the rock that Rambus came out from under, lift it up, and chuck them back.
________________________________________
Napster-to-go says "Fill and refill your compatible MP3 player", which is a lie. It's not MP3. It's WMA with DRM.
Don't count on these birds of woe disappearing anytime soon. Even if all of their patents are held invalid they are still going to collect a lot of money from people. Remember - they don't even have to advertise - which is a major expense for most companies. The advertising and promoting bills go to their 'clients'. What a scam these guys have going.
You never say "Johnny are going to the store", do you?
Okay, I'll bite...
Johnny is a person. Intel is a company. In some dialects of English, companies are referred to in the plural. It's no worse than the trend to use "they" as a generic third-person singular: "If someone steals my s33kr1t DivX movie collection I'll send my flying aardvarks after them!"
I just canceled my order for a Dell 620 which uses PIII Xeon and RDRAM (fortunately still sitting on the secretaries desk). I'm curious though. I've been searching and cannot find any benchmark numbers comparing a PIII Xeon to a regular PIII (e.g., otherwise identical Dell 620 v. 420). I find it rather amazing. Tom's hardware et al. seem only interested in 3D game performance. Where's the data? Has no one run SPEC CPU2000 on a Xeon? (Perhaps that says enough!)
The _idea_ behind rambus memory is a good one--
:-) for
there are some very impressive bits of technology
that could (and should) be developed.
HOWEVER; Rambus the _company_ has dropped the
ball. Rather than develop the technology, they
have behaved in the finest tradition for the year
2000, and tried to make money by whinging,
whining, and ultimately suing everyone doing
better than them. They behave like children, and
I've been gleefully watching them get spanked.
Maybe I'll form a company to create electronic
dictionaries. Rather than collecting any words
though, I'll just sue Oxford (and Webster, but
definitely not the American Heritage group
copyright/patent infringement. MUCH easier than
real work.
"People who do stupid things with hazardous materials often die." -- Jim Davidson on alt.folklore.urban
Case in point, some have little awareness of technology or the workers who produce it. Good reading can be found here.
Perhaps more to why many hold a dim view of Intel would be that not every slashdotter believes Intel makes the best technology. Many, myself included, have worked on other platforms (many of which are vastly superior designs) which have been marginalized by the dumping of cheep WinTel boxes.
--
Chief Frog Inspector
A feeling of having made the same mistake before: Deja Foobar
Wraithmaster
www.wraithmaster.com -- Chicken soup for the spleen.
www.wraithmaster.com -- Chicken soup for the spleen.
"Naaarf!" --Pinky
Why the hell is it so hard to make non-buggy products with Rambus? I first heard about Rambus when evaluating some new SGI graphics adapters (High Impact and Maximum Impact) that were among the first devices to use Rambus DRAM. Guess what? The products were announced in mid-1995 but ran into various bugs and supply problems which led them not to really ship till beginning-1996. When those supply problems became apparent, their stock hit an all-time high (~45) never to return and it has since dropped to around ~4.
I talked to a competitor (DEC)'s engineers around that time and they said that while they'd looked at Rambus, it was not a very stable memory technology; the complexities it introduced into their engineering were not worth the performance gain and cost hit.
The fact that Intel has had the same problems as SGI, albeit on a much, much larger scale, really leads me to wonder... what is it that makes Rambus memory controllers or interfacing chipsets so damn difficult to get working properly?
If someone could answer this question, I'd be really obliged. The "toll-keeper" problems were obvious from day 1 with Rambus... the techical problems were not.
--LP
First off, Intel have never OWNED Rambus - they merely license their technology for use in their various products. Anyone saying otherwise is simply talking rubbish.
However, they DID have a fairly serious financial deal that enabled them to purchase lots of Rambus stock at knock down prices once they had met certain requirements - this indicates more than a passing interest in the wellbeing of the Rambus company.
Personally I'm still not convinced that Rambus ever had that much to offer on the technology front, so Intel's claim that they were solely in it for the technology strikes me as doubtful. RDRAM has lots of extra bandwidth for sure, but when Intel forced the system on to the industry there was really no need for vast bandwidths such as these.
Moreover, RDRAM has really no electrical characterstics that are superior to those of SDRAM, and the gap in cost is considerable even today. Furthermore RDRAM isn't really any more 'future-proof' than SDRAM, especially when you consider the future potential of DDR.
So maybe Intel have finally just woken up and 'smelt the coffee', so to speak, realising that Rambus doesn't hold the answer to all the technical problems faced by the memory industry.
Quite apart from the technical problems CAUSED by Rambus - who remembers the i820 fiasco?
Lots of pro-Rambus zealots like to point at the success of the Playstation 2 (which employs RDRAM technology in favour of more traditional SDRAM or even SDRAM DDR technology) and intimate that this somehow relates to PCs and that Rambus is therefore "better". Saying that Rambus is "better" just because it is used in a certain product is ridiculous. Windows is fairly widely used but we all know that it's not necessarily better...
So to sum up:
1. Intel are wise to get out while they still can.
arnald
Companies being treated as collective plurals is the proper idiom in British English. They have a right to it; it is their lanugage, after all.
-- Anne Marie
I found this section amusing, "Then, more recently, Intel scrapped its delayed low-end Timna microprocessor after customers showed little interest in a low-cost PC microprocessor that required expensive Rambus memory chips."
PHB : Ok boys, we want to sell to the skinflint consumers who only buy cheap Acers at CompUSA. I think the best option is to go with a cheap CPU and really expensive memory! How's that sound?
Lackies 1-N : That's a good idea! Let's do that!!
PHB : Good answer. Now let's have a power lunch.
-----
D. Fischer
ShoutingMan.com
"In retrospect, it was a mistake to be dependent on a third party for a technology that
gates your performance"
I suppose this doesn't apply at all to the Apple-Motorola situation, does it?
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
"This release contains forward-looking statements regarding financial results for future periods. Actual results could differ materially. Among the factors which could cause results to differ materially is the possibility that the Pentium 4 and PlayStation2 ramps will be slower than expected, that shipment of Rambus ICs and other licensed products by Rambus licensees will be below forecast, that no additional licenses for SDRAM-compatible ICs will be signed, that prices of RDRAMs will remain high compared to SDRAMs and that litigation and building costs will exceed the Company's plans."
Sounds like Rambus might just take care of itself!