SI
SI
discoversearch

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 : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?


To: techreports who wrote (78858)10/1/2001 8:09:27 PM
From: Skeeter Bug  Respond to of 93625
 
>>So comptuers have not been commodities in the past? <<
not anywhere close to now. the prices of computers and components have collapsed. folks REFUSE to pay up for computers. $2k+ computers were common a few years ago. not now.

>>If computers are commodities, then you better tell Intel and Microsoft they are screwed<<

microsoft had it first DOWN earnings q in its history. they don't need me to tell them they are already screwed as their bubble pe falls in the grip of rational market gravity. intel has had it WORSE than msft.

>>Again, if Intel wants RDRAM as the standard, that's what people will buy<<

you are wrong. intel tried to dictate another memory standard and they LOST b/c the market drives intel, not the other way around. perhaps bilow will recall the name of the memory intel foisted on an unwanting market. this has happened before.

when you are proven wrong, please be man enough to admit it to me on this thread. IF rmbs dominates dram then i will provide you the same courtesy.

>>. Second, when did people give a flip about what kind of RAM is inside their computer (the majority of people don't care. Only nerds like yourself do.).<<

they don't. never said they do. what they care about IS THE PRICE OF THE COMPUTER. how you got off on such a tangent, i don't know.

>>And for your comment that RDRAM might not provide any increase in speed, then explain the reports I've read that a 1.5GHz P4 with RDRAM out does a 1.7GHz P4 with SDRAM, yet costs less?<<

please provide some links to these reports and the pricing. i've already said that under certain circumstances, rdram performs better. under others, i've read it doesn't. tell me, how much faster can i type this note with rdram? you see the tree, but not the forest.

>>again, RDRAM adds like 40-60 dollars to the cost of a computer. If you are buying a $2,000 dollar machine, it doesn't really matter. Granted, if you are buying a 600 dollar machine, then it probably does.<<

tech, please show me where the same company sells an rdram box for $40 more than a ddr box when each is configured as close as possible to the same. wrt $600 computers, guess what kind of computers make up the bulk of sales in 2001? not $1k+. so, we agree, it does matter. so does the market b/c they could buy rdram enabled boxes in droves... but aren't....

>>oh, i keep forgetting. People buy computers because of the type of ram inside them.<<

tech, i said "the price is right" not to highlight dram content but to highlight "the price is right." i understand we all have slow days, but did you get it this time?

>>Again, i think you are providing trash. At least Bilow offers some original reasons backed up with facts.<<

you are entitled to disregard basic economic decision making. it is, and has been, at your own risk. facts will follow the basic economics here. i predicted rmbs' troubles quite a few years ago, on this thread, based on a quick, albeit basic (basic was ALL that was needed), economic analysis. my view panned out. just like my declaration after micron filed their suit that said rmbs committed fraud at jedec - here on this thread. common sense is not common b/c it is not understood by all. especially folks who have a financial and emotional interest.

now, we all eagerly await the facts to back up your assertions...

1. show us all equally rdram boxes within $40 of equally configured ddr boxes.
2. show us all the information behind the tests that showed rdram outperformed ddr.



To: techreports who wrote (78858)10/1/2001 8:35:35 PM
From: Skeeter Bug  Respond to of 93625
 
hardwarezone.com

>>RAMBUS had an even bigger blow when Intel openly showed results with PC133 SDRAM matching the RAMBUS DRAM<<

intel is always right, right?

>>This caused users to begin questioning the performance of RDRAM and whether it was justified to pay such a hefty amount for that little gain in performance.<<

basic marginal benefit vs marginal cost economics. how simple...

in reality, i'm not so sure the market will catch on to ddr very quickly unless there is no price increase.

news.cnet.com

>>"You're going to take a performance impact," said Dean McCarron, principal analyst at Mercury Research. "The reality, though, is that it doesn't matter" for most corporate customers.<<

a couple points...

1. a 10% improvement is only on memory intensive tasks, which make up a small % of most users' tasks. the overall impact is about 1-2%.

2. my company could increase productivity more by purchasing motion sensor devices that shut down screen savers vs purchasing rdram. come on!



To: techreports who wrote (78858)10/1/2001 8:42:05 PM
From: Bilow  Read Replies (3) | Respond to of 93625
 
Hi techreports; Re: "And for your comment that RDRAM might not provide any increase in speed, then explain the reports I've read that a 1.5GHz P4 with RDRAM out does a 1.7GHz P4 with SDRAM, yet costs less?"

That an RDRAM equipped computer costs less than the equivalent SDRAM equipped computer is a Rambus lie. RDRAM is more expensive than SDRAM, RIMM modules are more expensive to manufacture than DIMM modules, and RDRAM motherboards are more expensive to manufacture than SDRAM motherboards. Enough said.

To compare expensive dual channel RDRAM systems with inexpensive single channel SDRAM systems isn't very fair. Instead you should compare expensive dual channel RDRAM systems with expensive dual channel DDR systems. But you won't have any such DDR systems to compare with until next year (other than the nVidia nForce 420).

The naked fact is that there are far more computers with inexpensive single channel memory systems sold computers with expensive dual channel memory systems. Single channel systems are mainstream, dual channel systems are high end.

Rambus had a single channel RDRAM system, it was called "Camino" and it was a famous disaster, with low performance, poor stability and high costs. You can now buy the motherboards for $42 on pricewatch. Intel had another single channel RDRAM system in design, it was called Timna, and it was cancelled for the same reasons that plagued Camino.

Now the word is that Intel has a dual channel DDR system. When it gets here, supposedly about a year from now, then you can compare expensive dual channel DDR P4 motherboards to expensive dual channel RDRAM P4 motherboards. Here's a prediction: The DDR will win.

In fact, if it wasn't obvious to Intel that RDRAM was going to stay expensive, and that DDR would be cheaper and better, Intel wouldn't be going to the trouble of designing these new DDR systems.

Intel got themselves into trouble by signing that contract with Rambus and now they are slowly getting themselves out of it.

Re: "RDRAM adds like 40-60 dollars to the cost of a computer. If you are buying a $2,000 dollar machine, it doesn't really matter. Granted, if you are buying a 600 dollar machine, then it probably does."

You're making up numbers here. If I'm going to buy a $2000 computer, I'm going to put 1GB of DRAM onto it, not the 128MB you're probably assuming. Four 256MB DDR PC2100 DIMMs costs $104 on Pricewatch right now, while the same amount of RDRAM RIMMs costs $328. That's a difference of $224. If you have the vendor supply your DRAM, the difference will be even greater. Your estimate of $40 to $60 would be the difference for some dinky $600 computer with minimum memory, not a high end machine intended for photoshop applications.

Yeah, I know that you're a Rambus moron, and it's impossible to argue with you about these things, but think about it. Why did Intel make the 845? Why is Intel supporting SDRAM for the P4? Why does PriceWatch now have 7 pages of hits for "845 SDRAM" and also 7 pages of hits for "850 RDRAM"? Why did Jdaasoc, a long time Rambus investor and pump, suddenly announce that SDRAM will be the main memory for the P4? Why is Dell supporting SDRAM for the P4? Do you think that Dell will support DDR for the P4?

You're way past the point where it was reasonable to suggest that RDRAM had a future. It's time to admit that Rambus' only chance is in the legal arena.

Re: "This is also ignoring the fact that RDRAM prices will come down in price. What happens if RDRAM is as cheap as DDR?" That's easy. If RDRAM is as cheap as DDR, monkeys will fly out of my ass. Since I haven't had a monkey / ass problem recently, I'd say that it's not going to happen. And Intel said the same thing:

Paul Otellini (Intel VP) on RDRAM:
"The problem is the cost structure. The cost structure never met the goals we had."
news.cnet.com

What part of "the cost structure never met the goals we had" do you have trouble understanding? What happened with RDRAM is that the industry had trouble making the stuff, and told all the users (memory designers like me) that DDR would be the next mainstream memory. They promised us DDR at the same price as SDRAM, and most of them never even bothered to put RDRAM into production. Meanwhile the RDRAM makers (Toshiba and Samsung) never promised us that RDRAM would be as cheap as SDRAM, and wouldn't make any unless they had the orders in hand. To design engineers that spells "niche memory", and specifying niche memories is how hardware design engineers end up being forced to write software (or drive taxis) for a living, so we designed in DDR instead. Only a few companies went with DDR, and it was due to their management making the mistake of ordering their engineers to use it or walk. The companies that designed in RDRAM recognized their error and used DDR or SDRAM for their next products (the one you've heard the most about is Nintendo). New companies showing up with no baggage chose DDR over RDRAM (Microsoft, nVidia, ALi, SiS, VIA, ATI, Cisco, IBM, Agere, AMD, Broadcom, PMC-Sierra, Transmeta, ST-Microelectronics, etc.) because DDR was widely supported by lots of industry stuff like the FPGAs available from Xilinx and Altera, while RDRAM was hardly supported at all. This convinced us to use DDR instead of RDRAM, and that made DDR the standard. The memory makers left RDRAM as a niche memory to die on the vine. They never manufactured it for the spot market, never gave pricing on it comparable to SDRAM, and just to make sure that RDRAM never got out of niche status, they further broke the RDRAM chip market into 6 different sub markets each with incompatible pinouts. By comparison, JEDEC standard DDR comes in only 3 pinouts (x4, x8, x16), and those three pinouts differ only in the width of the data bus. Get over it, the RDRAM story is over.

For a description of why design engineers chose DDR over RDRAM see: #reply-15751210 The thing to remember is that there is a delay of about 1 to 3 years from when design engineers decide on what memory to use in a chip to when actual products using that chip are available for sale. The engineering decisions as to what memory types will be used in 2002 were decided in 2000, and at that time RDRAM was horribly expensive. The fact that RDRAM has dropped in price is great, but it's too late for the RDRAM designs that instead used DDR last year. And DDR is still a lot cheaper than RDRAM, so none of those engineers who chose DDR is having any problems keeping his job (at least for that reason).

-- Carl



To: techreports who wrote (78858)10/1/2001 9:02:58 PM
From: Skeeter Bug  Respond to of 93625
 
tech-report.com

>>Finally, after sticking with its memory partner for quite a while, Intel saw fit to begin distancing itself from Rambus late last year. The two companies still work together, but Intel has slowly and deliberately begun pursuing an alternate course by making SDRAM chipsets available for the Pentium 4. Since then, much of the market has been waiting for the Pentium 4 to transition to a new, 478-pin socket—and, inevitably, away from RDRAM.<<

take a moment and flip through the various tests provided here (the verbage doesn't always appear to match the graphs, which is strange).

conclusion?

tech-report.com

>>For all the brouhaha over DDR versus Rambus, the real-world performance difference between the two technologies—at least with these chipsets—is statistically insignificant. Nonetheless, VIA's achievement with the P4X266 is striking. They've achieved performance parity with RDRAM, and with Intel's own native platform chipset<<

>>Given how well the 845 extracts bandwidth from PC133 SDRAM, we can't wait to see it paired up with PC2100 DDR memory when the time comes. Nevertheless, the 845 is poised for sales success now. It enables the Pentium 4 to take its place on corporate desktops [ed, whereas it CAN'T take its place with rmbs], where performance isn't paramount [ed, rmbs cost / benefit ratio sucks!], and competition from AMD and VIA isn't likely to present a problem.<<

tr, in the real world, people and companies don't 1. take risks with new technology and 2. pay significantly more money for statistically insignificant performance - their words, not mine.

economics contains some powerful precepts. ignore them at your own folly.