Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: wanna_bmw who wrote (73785)	3/6/2002 10:51:51 PM
From: Joe NYC	Read Replies (2) \| Respond to of 275872

wbmw, Link please. I think I asked you first. The right answer is not to ask the other person for his link but to deliver. Here is my link: siliconstrategies.com Intel commits to using only Rambus with upcoming Tehama chip set By Will Wade EE Times (05/09/00 18:15 p.m. EST) FOLSOM, Calif. ( ChipWire) -- Intel Corp. is running prototypes in the lab of core logic for its upcoming Willamette microprocessor, featuring support for Direct Rambus DRAM technology. Both the core logic and Willamette are due for commercial release by the end of the year. But analysts are concerned that Intel's focus on RDRAM could backfire if the memory chips are not available at competitive prices by then. Intel has working silicon for both Tehama, the primary chip set for desktop PCs using the Willamette processor, and Colusa, the chip set designed for the server version of the Willamette MPU, code-named Foster. Reiterating Intel's dedication to RDRAM, Louis Burns, vice president and general manager of the company's chip set group, said, "From a technical point of view, RDRAM is absolutely the best solution." While PC OEMs have started to ask memory vendors to provide double-data-rate (DDR) SDRAM to increase system performance at prices lower than for RDRAM, Intel's new processor will only work with RDRAM. "My roadmap does not include any DDR," said Burns. However, Intel had said the same thing a year ago on the eve of the release of its 820 chip set, code named Camino. That chip set was designed to link the Pentium III processor with memory, and Intel insisted that was its only function. Since then, RDRAM chips have been in short supply and system OEMs have been reluctant to adopt the expensive technology in volume. A few months ago, Intel announced that the 820 also works with a memory translation hub (MTH) that allows the chip set to run with standard SDRAM (see Feb. 18 story). Dean McCarron, principal analyst for Mercury Research in Scottsdale, Ariz., said that about three-quarters of the Camino chip sets shipped in the first quarter of this year were intended to work with SDRAM. "A lot of people are using the MTH," he said. While Burns insisted that Intel has no plans to produce anything similar to the MTH for the Willamette and Tehama product generation, McCarron said it would not be difficult for the company to quickly produce such a chip if necessary. "Intel is fanatically committed to Rambus," said McCarron. With a top speed in the 800-MHz range, RDRAM is much faster than the 266 MHz seen with current samples of DDR SDRAM. However, McCarron said, RDRAM has a longer latency period than SDRAM-based designs. And without full systems using both technologies for a side-by-side comparison, it is still difficult to name one technology as the clear performance leader. Performance may be moot, though, because cost is one of the most important remaining questions. RDRAM is at least three times the price of SDRAM, said McCarron. This is a function of the still-limited availability of RDRAM, as well as their lower yields compared to SDRAM and the mandatory royalty fees that must be paid to Rambus Inc. DDR SDRAM is projected to be almost the same price as SDRAM by the end of the year. "RDRAM availability is incredibly important for the Tehama and Willamette launch," said Intel's Burns. Your assertion was that: I don't think there was anything to hide. I think Intel announced long before the i845 with DDR memory was launched that they wanted to do a DDR chipset, Which is not the case. Long before i845 was launched, Intel was saying Rambus was the only memory type supported, while they were secretly working on SDR and DDR chipsets. Then came the silence: Meanwhile, the issue of if and when Intel will permit third parties to support its CPU with chip sets built for alternative SDRAM and double-data-rate (DDR) memories still goes unanswered. siliconstrategies.com Notice, it is a direct contradiction of your assertion And suddenly after 6 months of lying, 2 months of silence, (8 in total), chipset samples arrive in Taiwan: Intel's Brookdale chipset arrives in Taiwan siliconstrategies.com No announcement that Intel wanted they do a DDR chipset prior to 845. 845 appeared out of nowhere, and only when it arrived, Intel said they wanted to DDR, after it was obvious that their plan to corner the market with Rambus failed. Joe

To: wanna_bmw who wrote (73785)	3/6/2002 11:24:13 PM
From: Joe NYC	Read Replies (1) \| Respond to of 275872

wbmw, "Withholding" implies that Intel had chipsets that could support the memory, but they were unwilling to launch them. That's silly, and you know it. Clearly, Intel made the commitment to go Rambus, and at the time, they had no hedge plan. Therefore, they had to scramble to make solutions when RDRAM wasn't competitive. Right now, RDRAM is very competitive, but it's too late. The industry wants nothing to do with it. When industry standard memory is PC-133 SDRAM, PC-1600 and PC-2100 DDR, and Intel refuses to support these industry standard memory types, and instead supports a proprietary one in order to corner the industry, to me it equals withholding support. Re: "I wonder if 2 to 3 years from now you will also agree than Intel is making the right decision by going to x86-64, abandoning than IA-64?" If that comes to pass, then why wouldn't I admit that I am wrong? LOL. You are doing it again. You are already planning on flip-flopping your opinion to be in tune with Intel agenda. If you really were an independent thinker, you would have said: "If Intel goes with x86-64 and kills IA-64, it would be the dumbest idea, because IA-64 is far superior to x86-64 because of a), b), c) (insert your PR gibberish such as robustness, high availability etc) The examples of why you are an independent thinker that you cite: Most recently, I have been critical of Intel's megahertz myth, which they seem to be propagating when they now have the chance to do something about it. AMD hasn't been able to get TPI off the ground, and instead they are relying on QuantiSpeed. And you knock QuantiSpeed, since Santa Clara has just given orders. I can't believe how you are playing into what TWY and Kap have been saying. Re: "I have been AMD shareholder on this board, and I have questioned probably half of the decisions AMD made. You seem to be content with every single decision Intel makes, until Intel makes a reversal, and then you support the new decision, keeping you 100% in sync with Intel PR operation." Come now, Joe. You shout the Jerry Sanders Party Line more often than not. Are you so blind that you can call me 100% biased, and you 0%? Common, be real. You can go through my post and find 100s if not 1,000 posts of various criticism of AMD, disagreement with their strategy and execution. The best you could do is to give a neutral comment about MHz propagating the MHz myth (BTW, wasn't it one of the speaches at fall IDF - moving beyond MHz) and you did not forget to sneak in a shot at AMD, while you were supposedly demonstrating your free thinking. Re: "If you think this is an unfair portrayal of you, maybe you can list instances of current Intel policies, actions, direction or decisions that you just absolutely hate or even things that upset you or you are annoyed by." I already have, and if you stay tuned, you're sure to hear more. Ok, I will be watching. Joe