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.

Strategies & Market Trends : Gorilla and King Portfolio Candidates -- Ignore unavailable to you. Want to Upgrade?

To: chaz who wrote (7241)	9/30/1999 2:45:00 AM
From: Bruce Brown	Read Replies (1) \| Respond to of 54805

It's not quite on your point, but we have to keep in mind that, as far as we know now, the glitch is at Intel's house, and that it means revenue deferral, not revenue elimination. If a competitive product was running neck-and-neck against RMBS and a delay like this came up, it would be very serious. But there's not a competitor. RMBS still has next generation memory to itself. It appears that the Camino/motherboard glitch is not the only glitch coming out of Intel these days. dailynews.yahoo.com In brief: ---------------- SAN FRANCISCO (Reuters) - Intel Corp. (Nasdaq:INTC - news) said it has discovered a bug in two versions of its Pentium III Xeon processors for the server and workstation market, a bug that will delay the shipment of servers based on the chip. Intel said it is still shipping the chips -- a Pentium III Xeon with a speed of 550 megahertz and a level two cache of 512 kilobytes, and the other with one megabyte of secondary cache. Both chips run in configuration of eight processors on an Intel motherboard, called the Sabre. Cache is a reserved section of memory to improve performance. The Pentium III Xeon 550 megahertz with two megabytes of level two cache is not experiencing any problems, Intel said. ``We expect to have the root cause and solution within the next few weeks,' said Chuck Mulloy, an Intel spokesman. ``We are still shipping those, but telling customers not to sell those systems and we are not shipping them with the Sabre motherboards right now.' ------------ To put the Rambus/Camino delay in perspective along with the Pentium III Xeon/Cache and previous product announcements over the years, we need to be reminded that smooth adoption (as if adoption is ever smooth!) of a next step technology is more difficult than we sometimes give credence. It's part of the process and amazing how the noise and FUD level has increased over the years when it comes to delays and glitches. Even at Apple, which is more of an in house operation because the OS/hardware union softens the coordination somewhat, experiences many glitches and delays. The PC industry with all the OEMs, Intel, Microsoft and logistics of coordinating the entire circus amazes me that progress can ever be made on an announced schedule without major glitches. I wouldn't be so quick to throw the towel in yet on Rambus, Intel and the OEMs coordination. I think they are simply experiencing the normal process that must be gone through to arrive at the next step wave of product. Having so many good cooks in the kitchen at the same time is bound to create an accepted level of chaos, but the chances of experiencing the collective effort meal which has never before appeared on the menu - once the damn thing is finished - is mouth watering to say the least. We have learned through our GG process that a degree of safety comes from paying for the meal once it has been eaten - or is at least is in the take out container for those who are always on the run. Some of us, myself included, pre-paid for our Camino/Rambus meal with the assumption that it would taste so good we overlooked the fact that until the smoke stops billowing out of the kitchen - let the cooks get it together no matter how long it takes before salivating. The appropriate time to drool will come. BB

To: chaz who wrote (7241)	9/30/1999 3:27:00 AM
From: FLSTF97	Read Replies (2) \| Respond to of 54805

As to Rambus, my thought is that its Gorilla potential is still intact. I propose that (only partly facetiously) that Intel is dependent on Rambus. Assume that Intel makes the world's fastest CPU: 10 x that anybody else. Unfortunately in the real world it only executes programs 1% faster than the competitions' because the bottleneck becomes the memory. That means that their once prized, highly differentiated, high margin products slips closer to being commodity products. This is certainly not what Intel plans on letting happen. You can bet that Intel did not feel compelled to "share" $1 bil (or was it $700 mil, I forget) with Micron and Samsung because they had an overwhelming sense of altruism. As to RAMBUS being hard to make work...big deal. I remember when people were adamant that MOS devices would never work, but then they learned to control sodium contamination. I also remember people then being certain that CMOS would never amount to much but then they started to understand latchup problems. The fact of the matter is that no IC or high frequency mother board is easy to build, but rest assured that they will be built. Will that make them a Gorilla? Maybe not. Not every memory application will need the speed advantage to justify the additional cost. Is it nevertheless a good investment? That depends on how much of the market you believe they will really capture. I personally think it will be less than 50% by 2003, BWDIK! Fatboy

To: chaz who wrote (7241)	10/1/1999 5:26:00 AM
From: unclewest	Read Replies (1) \| Respond to of 54805

Not the case. By Uncle West's measure, 65% of the Rambus revenue is non-Intel, and it's growing, other partners are dedicated to RDRAM. Intel would grow the revenue faster to be sure. It's not quite on your point, but we have to keep in mind that, as far as we know now, the glitch is at Intel's house, and that it means revenue deferral, not revenue elimination. If a competitive product was running neck-and-neck against RMBS and a delay like this came up, it would be very serious. But there's not a competitor. RMBS still has next generation memory to itself. chaz, just to clarify... the 35/65 figures came from rmbs cfo at the last shareholders meeting (feb) which i attended. the revenue model he presented called for 35% from pc's, 15% from servers and 50% from other. we know the pc status. the server chipset is still under development by intel, scheduled for sampling this year. the other 50% is more difficult to pin down, but "other" product announcements have been forthcoming. many are kept secret for business reasons. i am most troubled presently by the latest samsung announcement to stop producing wafers for rdram. it seems samsung is not satisfied with whatever word they are getting from intel re. camino 820. interesting to note that nec announced increasing rdram production last week. rmbs share price makes huge swings based on the market's perception of rdram in pc's...rdram in "other" is more difficult to see, therefore it is largely ignored by investors. jmho, unclewest i am a gmst player also edit...with $ billions already invested by over 100 companies no expense will be spared to find a solution.