Before you build a P4 read this!

[chrisojeda]

This week Rambus faces Fraud charges. Before building a P4 I would state that you should consider the following (below). I am not saying don't do it. I am just stating that you should consider whether under the circumstances it is the best option.

It is likely that Rambus will be found guilty of Fraud charges. Every single motion against Infineon. If Rambus is found guilty you can expect them to pay hefty penatlies and possibly be penalized in a non-momentary way. This will surely bring down the stock value hard. Even if they are found innocent their stock will more than likely suffer for some time.

The companies that paid royalties to Rambus will now be taking Rambus to court to get their money back ("with interest"). This is going to be yet another financial blow. This one is going to cripple the company.

Rambus is in big trouble financially for the above two reasons but also all of the litigation is costing Rambus millions. The ace up their sleeve was to win the case and thus become paper rich.

Intel may finally have a way out of having to use Rambus technology. I can't imagine the contracts between Intel and Rambus not giving Intel a way out if Rambus should either fail to produce the required quantities (in-house or via licensing) or have any negative image in the public. Considering all the financials troubles Rambus may not be able to deliver product. Considering the Fraud case, Rambus's bad reputation may negate the contract if that term was stipulated (which almost always it is).

I think one should give serious thought before investing in anything that uses Rambus RAM. It just might be a dead end system with no support in the future.

That's my $.02.

It worked yesterday! :lol:

 

[Guest]

Good Post!

The royalty that RAMBUS charges for DDR-RAM is 3.5% of sales. The market is expected to hit $50 billion a year. Can you do the math on this? Just astpunding what a toll this could have been!

 

[Raystonn]

"if Rambus should either fail to produce the required quantities"

Rambus doesn't make any RAM at all. They were just the original designers of RDRAM. Dozens of memory manufacturers are the ones responsible for creating RDRAM. Should Rambus go extinct, RDRAM will go on. It would then become an open standard.

"It just might be a dead end system with no support in the future"

It has quite a future actually, completely independant of that of Rambus as a company. RDRAM is really taking off in the industry, with dozens of memory manufacturers creating it. Currently there is more demand than supply, so we can expect the industry to keep ramping up production, and prices to continue dropping.

I recommend performance and price to be the main considerations when deciding on what type of memory you want for your system. RDRAM is definately not going away any time soon.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[AeroSnoop]

Not meant to be a flame or anything. I was just wondering, if Rambus did really have to pack up and go would Rdram really become an open standard? It seems to me like Rambus would still want some sort of royalty for rdram itself.

 

[Raystonn]

"if Rambus did really have to pack up and go would Rdram really become an open standard? It seems to me like Rambus would still want some sort of royalty for rdram itself."

Well, either Rambus exists or it does not. If it exists, then there's no problem with RDRAM's future remaining on par with what it is today. If they do not exist, they cannot ask for royalties. In this case, the company that holds the patents/copyrights no longer exists and, unless these patents are sold, they fall into the public domain. If the patents are sold, the purchasing company, perhaps Samsung, would then be collecting the royalties and designing the standards for RDRAM.

In any case, RDRAM still continues to thrive as a technology. Technology is independant from companies.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[AeroSnoop]

Thanks, I wasn't sure what happened to patents when a company goes under.

 

[Guest]

Good Post!

The royalty that RAMBUS charges for DDR-RAM is 3.5% of sales. The market is expected to hit $50 billion a year. Can you do the math on this? Just astounding what a toll this could have been!

 

[Raystonn]

This post was about RDRAM, not DDR.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[TheAntipop]

raystonn, do you have large quantities of Rambus stock or something because you are the only one to every stick up for the technology.

<A HREF="http://static.stileproject.com/pika.swf" target="_new">Hyakugojyuuichi!!</A>

 

[TheAntipop]

and on a separate note, would you personally even consider buying those cheap RDRAM sticks that go for the low end of the prices? many, many people have learned their lesson in buying cheap products because of attractive price, so does the lowest price for the memory actually tell us anything if few intelligible people are even going to consider it?

<A HREF="http://static.stileproject.com/pika.swf" target="_new">Hyakugojyuuichi!!</A>

 

[Raystonn]

RDRAM goes through strenuous testing before it's approved for sale. This is actually one of the things that makes it slightly more expensive to make. SDRAM doesn't have this because it's an open standard. Noone wil threaten to revoke an SDRAM license if a manufacturer releases subpar components. You can basically be assured that if it's listed as PC800, it really is. It makes them a bit more expensive to produce, but I like the extra quality.

Just to answer yoru previous post, no I do not own any Rambus stock. I personally don't like the actions Rambus's lawyers have taken. (Remember I don't necessarily speak for my employer here.) But I do like RDRAM as a technology.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Guest]

Surely this is good news! If RAMBUS looses its patents and royalties, it will open the market. We could be looking at cheaper, faster RAMBUS very, very soon. Plus, the Taiwanise would be fools to change the format just because the companie dissolves.

~ The First Formally Rehabilitated AMD Lemming ~

 

[beans]

Raystonn -

You say, "RDRAM goes through strenuous testing before it's approved for sale. This is actually one of the things that makes it slightly more expensive to make."

Do you think that, should Rambus' patents be invalidated, RDRAM could be made at a cost that would allow it to compete on price alone?


beans

 

[Raystonn]

"Do you think that, should Rambus' patents be invalidated, RDRAM could be made at a cost that would allow it to compete on price alone?"

Rambus's RDRAM patent will not be invalidated. The court cases are about SDRAM, not RDRAM. Eventually RDRAM will compete based on performance alone because price will be on par with other memory technologies. This same thing happened when SDRAM was first introduced into a market full of 72-pin RAM. SDRAM was extremely expensive at first due to low supplies and high demand.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[beans]

Raystonn -


Do you mean that RDRAM's cost of manufacture is, or can be, as low as that of SDRAM?

beans

 

[Raystonn]

"Do you mean that RDRAM's cost of manufacture is, or can be, as low as that of SDRAM?"

Once production has been ramped up and the overhead of buying the new testing equipment has already been absorbed, yes, RDRAM will cost about the same as SDRAM to produce. The situation we are now seeing is exactly the same as that when SDRAM was introduced to replace 72-pin RAM.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[beans]

Raystonn -

You mean there's a shortage of RDRAM relative to the demand for it, and that's what's driving the price up?

beans

 

[Raystonn]

"You mean there's a shortage of RDRAM relative to the demand for it, and that's what's driving the price up?"

The price for RDRAM is not going up, but down. The main factors that cause RDRAM to be expensive to begin with include the requirement for new testing and manufacturing equipment and slow production speeds. The more RDRAM a manufacturer pumps out, the less each module costs. Also, the slower production moves, the less modules are are out the market. SDRAM is very cheap right now because the market is flooded with it. This will eventually happen to RDRAM as well.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Guest]

Chrisojeda wrote:

"It is likely that Rambus will be found guilty of Fraud charges. Every single motion against Infineon. If Rambus is found guilty you can expect them to pay hefty penatlies and possibly be penalized in a non-momentary way. This will surely bring down the stock value hard. Even if they are found innocent their stock will more than likely suffer for some time."

You are absolutely right. Take a look at this:

http://www.cnbc.com/chart/chart.html?sym=RMBS

It's down 8.5% in one day!

 

[beans]

Raystonn -

What I meant by "driving the price up" was "keeping RDRAM prices high with respect to those of other technologies."

In other words, are you saying all available RDRAM production is selling at current prices?


beans

 

[Raystonn]

"all available RDRAM production is selling at current prices?"

I do not understand what you mean. Everything in existence is 'selling at current prices.' That is the definition of 'current prices.'

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[beans]

Raystonn -


I guess I shouldn't post late at night.


At any given price, a certain amount of RDRAM will sell. There is a price at which RDRAM makers will sell everything they can produce.

If they set a higher price, less RDRAM will sell. Makers will idle some of their production capacity, and/or RDRAM will start to accumulate in their warehouses.

What I'm asking is, do you think RDRAM makers are setting prices low enough to sell everything they can produce, using all available production capacity? Or, are they setting a higher price and selling less than that?


beans

 

[Raystonn]

I'm sure they're selling every module they make. I'm not a marketing guy and I don't work with those manufacturers, so I really don't know for certain. The main factor keeping prices higher is they are not yet able to make enough. The manufacturers are continuing to ramp up production every day. The more they make, the less each will cost.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[SammyBoy]

I'm sure they're selling every module they make. I'm not a marketing guy and I don't work with those manufacturers, so I really don't know for certain. The main factor keeping prices higher is they are not yet able to make enough. The manufacturers are continuing to ramp up production every day. The more they make, the less each will cost.

Isn't it more or less true that most of the demand is coming from the OEM sector as opposed to the public sector? If that is the case, isn't it likely that once a cheaper P4 solution comes out (SDRAM or DDR) that most of the P4 systems built will use those cheaper components? That means, if true, that the public sector won't adopt it as much, and the demand will plummet. Also, since only P4 systems use it, there can't be that much demand right now, and I figured that most manufactures would have banked on the idea that since it's Intel, that it will sell well.

Seriously, I don't see RDRAM getting anywhere. When you look at the big picture, SDRAM is the king of the hill, be it PC100 or PC133. DDR-RAM is the logical successor to SDRAM, as it's the same type of RAM, just with an extra transmission in a clock cycle. Plus, DDR will eventually ramp up to 300 and 333MHz (PC2400 and PC2700) by the end of this year. Also, since AMD refuses to use RDRAM, a major CPU builder doesn't accept it, and since Intel will be building a DDR solution for the P4, and ones already exist for the P3, it is safe to say that a super-majority is behind DDR, and only a few are clinging to RDRAM. Don't forget that graphics cards use DDR memory and not a one has used, or been announced to use, RDRAM.

I think two things happened. First, since RDRAM has to be installed in pairs, it harkens back to the old days, where if you wanted to upgrade your RAM, it was tedious, and if you had 16MB RAM, and wanted 32MB, you either had to get two 16MB sticks, or two more 8MB sticks. If you do the former, you throw out two sticks, and if you do the latter, you increase the memory latency. It was expensive to upgrade, and too much hassle. Then SDRAM came out and the world rejoiced. No more pair installations. Want and extra 32MB? Don't need to buy two 16MB sticks, just get a 32MB stick. And guess what, the latency of the RAM isn't increased. The same goes for DDR. Second, I think that Rambus, in its haste to become a solid leader in RAM design, pushed too hard, and really pissed some people off. They screwed over a lot of other members on the standards board, and then went ahead with a technology that other groups had rejected/didn't think was fesible. The only thing that protected Rambus was its contract with Intel. At least then, Rambus was assured some kind of income and royalties and licensing fees. But alas, Rambus's past came back to haunt it, and Infinion (sp?) fought back. According to the news today, they not only kicked Rambus's suit out the door, but recieved punitive damages as well. I'm sure that Intel's legal department is scrambling to a) see if there is a way out of their contract with Rambus, and b) purge any traces of possible colusion (sp?) with Rambus. The last thing that Intel wants or needs right now is to be dragged into a legal battle of Rambus vs. the memory makers of the world.

Sorry that went on, but I really don't like the RDRAM technology, or the company that produced it. Plus, while it seems to help the P4, I really don't see it as being enough of a performance boost to justify any size of premium on SDRAM and DDR-RAM. Guess we'll have to wait and see if the DDR solution for the P4 will be on par with the RDRAM solution. Also, I would like to say that some people out there believe that the DDR solution to the P4 may be limited to PC1600, since Intel chips require memory to be at the same speed as the FSB.

-SammyBoy

Without Evil, there can be no Good. Therefore, without an Intel, there can be no AMD.

 

[Raystonn]

"since only P4 systems use it, there can't be that much demand right now"

Actually, there is a very high demand right now compared to supply. Intel sold its 1 millionth Pentium 4 during Q1 of this year and sales keep increasing. As far as the negative comments about the Rambus company, I will keep out of that. Regardless of Rambus, RDRAM is a great technology.

-Raystonn

= The views stated herein are my personal views, and not necessarily the views of my employer. =