Intel is not a party in this lawsuit. But don't be misled over statements as to "what Intel said."
Here is what Intel said publicly. Please note all the "myths" they do away with.
You make very wild and unsubstatiated speculation that Intel would lie to their Intel Forum members and the world about their commitment to Rambus.
Intel was ushering in a new and very promising technology (Rambus). The market was free to accept or reject that technology. The market (Dell/Compaq) accepted the technology and wanted it in volume so as to decrease the cost.
The cartel members (Hynix/Micron) were able to control both cost and availability of RDRAM and competing technologies, and did so illegally according to admissions to the DOJ. Why would they commit felony price fixing if RDRAM was so flawed? Because they wanted to control the market and not be reduced to "foundry status" as the emails clearly indicate.
This is not about technology. It is about felony price fixing, joint boycott, etc. Collusion. Conspiracy. Power. Call it what you like, but the jury will come back with some very large damage numbers very soon. Maybe today.
Of course execs were promoting RDRAM in 2000, they were already committed 100%, what did you expect them to say?
RDRAM was not the best technology available, but it offered higher bandwidths so intel made a huge gamble in promoting RDRAM because it would hopefully show off capabilities of Netburst architecture in P4 procs, but they didn't count on latency sucking so bad. Maybe they figured Rambutt would solve the issues before Intel was forced to drop them.
Intel lost billions of dollars and also lost a lot of market share to AMD, not to mention suffered a tarnished reputation as a result of this relationship.
remember the fiasco when Intel introduced rdram - and had to recall all the motherboards? here's a nice factual article .....
"I would argue that engineers know better than execs when it comes to the technology, so quit marginalizing them as "low level", it's demeaning.
Draw whatever inferences you want by "low level", but fact of the matter is that a couple of these engineers do not constitute what "Intel says." What "Intel says" is characterized by press releases and the the comments of top level executives, Pat Gelsinger being one. Gelsinger went on to say that Intel chose Rambus because "it was simply the best technology available.
As to the technical merits of RDRAM v SDRAM, that is your opinion and really doesn't matter. The issue was - what were Intel's intentions. From the highest levels of Intel, as echoed at the Intel Developer Forum, Intel chose Rambus.
BTW, the bandwidth comments were not mine. They were Mr. Gelsinger's (Intel). As he is now President and CEO at EMC, I would place more credibility in his comments than yours.
Meyer Services, 38 Lipsky St. Jerusalem Israel
"rdram.pdf" - First posted 5 October 2001.
/Producer (Acrobat Distiller 4.05 for Macintosh)
23 September 2001.
"As volumes ramp, 4i technology is introduced, and memory densities continue to increase,
RDRAM costs will continue to decline."
---Costs didn't particularly decline but pricess fell due to excess inventory and lack of interest due to lack of benefit. Weeks later in October 2001:
"Analysts told Electronic News Online that Santa Clara, Calif.-based Intel (nasdaq: INTC) has altered its road map, eliminating a P4 chipset supporting 4-bank RDRAM planned for mid-2002, code-named Tulloch."
"Dual-channel RDRAM provides 3.2 GB/second of memory bandwidth to the system."
--The alleged "myth" was that RDRAM had little performance benefit in many applications. The alleged "reality" ignores that and babbles about bandwidth. In fact a much less expensive Athlon and single channel DDR gave equal or better performance in many applications and Intel knew it.
Allegedly to dispel the "myth" that memory makers weren't making RDRAM (they'd already made almost enough to ensure that they'd never get rid of it all) : “Japanese chip and PC maker Toshiba* Corp said on Friday it will more than
triple its output of Rambus DRAM chips by September while cutting commodity DRAM production, and NEC
Corp said it is eyeing a similar move.” Yahoo Finance, 2/9/2001
--Y! FINANCE????!???!!!!11 How "authoritative". In December of the same year Toshiba had sold its U.S. DRAM and Flash ROM facilities to Micron (who didn't shut them down) and exited the commodity DRAM market.
"RDRAM systems are inherently more expensive to build than DDR systems"
--Not a myth, didn't need "dispelling".
"RDRAM will not be available for upgrades in 2-3 years"
--You call that an "upgrade"?
"Intel does not collect royalties on RDRAM memory technology."
--No, Rambus collected the royalties, they were just supposed to pass anything more than 2% through to Intel. Testimony under oath in San Fransisco has told us that Rambus collected more than 2% but didn't pass anything on to Intel. Double "myth", doubly "debunked".
"As soon as Intel or others launch platforms supporting SDRAM or DDR, Intel will drop RDRAM"
--They sure did. Like a hot potato with leprosy and drug resistant gonorrhea.
If anyone bought any RDRAM based Pentium systems on the basis of that misleading (if not outright false) advertising (and if it indeed was produced by Intel) then it's a shame. If people sensibly stayed away and Intel lost business and reputation because of such nonsense, all the better. It seems like this was more a production of Rambus, parroting their earlier "dispelling myths" BS that they paid anandtech to "interview" them for.
No to Rambus, and No to Intel domination!
Notice it was samsung who said this..“We think RDRAM will be very
price-competitive with DDR
Geoff Hughes, Samsung* senior vice president of
sales and marketing as quoted in EETimes, 3/29/01
Samsung is not part of the so called cartel, but they are the only ones who could control the market and price of both ddram and sdram..
When Rambus made peace with the king, Samsung, they lost the ability to prosecute the peasants.
Samsung was part of the cartel. They plead guilty to price fixing memory just like the rest of the cartel.
They are not in this case because they were smart enough to settle knowing the overwhelming evidence that Rambus has.
Samsung was smart enough to settle and leave the whole joint and several liability to Hynix and Micron, who were not so smart. The California "last man standing" rule means that Hynix and Micron will absorb the damages that would have accrued to Samsung. Not a great business move for the Hynix/Micron cartel.
I believe that Intel has said this was said publically due to arbitration in order to get out of the guillotine contract, and that privately they had 2 roadmaps, including DDR.
So you are saying that Intel has admitted lying to the world in 2001 when this came out, and specifically to the multitudes of technology companies attending the 2001 Intel Forum for whom this document was created.
Please reproduce a quote from a high-level Intel executive to corroborate your contention. You can't, and I don't think the jury would buy that explanation either.
Intel intended Rambus for it's mainstream memory for several generations of their chipsets as this document indicates.