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: Mathemagician who wrote (67206)3/6/2001 1:15:57 PM
From: Don Green  Respond to of 93625
 
Micron Hunkers Down
By Arik Hesseldahl

FORBES

The computer memory business has never been for the faint of heart.

It's an industry that never seems to attain a happy medium. One year there's a supply shortage that drives prices and profits skyward. Just as suddenly, supplies can stack up, driving prices down.

With such a roller-coaster cycle, it's no surprise that the number of companies making DRAM chips, or dynamic random access memory--the main memory chips used in personal computers, notebooks and servers--has declined to 14 from 28 since 1996. That number is only going to get smaller, and soon. The question for many chip investors is what happens to Idaho-based Micron (NYSE: MU - news).

It's not hard to guess the source of the market's current troubles. Slowing demand for PCs among consumers and business customers alike, coupled with the general slowing of the economy, has dashed hopes for a prolonged industry boom. As PC demand fell, supplies of memory chips have piled up.

In its most recent quarter Micron reported a $352 million profit on sales of $1.8 billion, or 58 cents a share, while analysts had been hoping for 60 cents. Among the analysts who have been cutting their earnings expectations for Micron in recent weeks is Joe Osha at Merrill Lynch, who trimmed his 2001 earnings outlook on Micron to $1.42 per share from $2.86 per share, citing inventory concerns.

Admittedly, the picture doesn't look good. Last year, worldwide DRAM revenue surged 40% to $28.9 billion. But this year the market is expected to decline by more than 11% to $25.7 billion, according to Semico Research of Scottsdale, Ariz., and that's one of the more conservative estimates out there.

Further consolidation in the DRAM industry is a near certainty. Building a new chip factory, or ``fab,'' is now a $2 billion undertaking. And shifts toward new manufacturing technologies and new types of memory require more expensive manufacturing machinery. ``It takes a lot of capital to stay in this business,'' says Sherry Garber, an analyst with Semico Research.

And a supply room stocked with antacids. As the slump persists, some companies may bow out of the business completely, or form joint ventures with others. Just last year, two Japanese companies, NEC (Nasdaq: NIPNY - news) and Hitachi (NYSE: HIT - news), soldered their respective memory divisions together to form Elpida Memory, which now ranks among the top five in the industry.

``There might only be four or five major players in the industry within a few years,'' says analyst Brian Matas of research firm IC Insights in Scottsdale, Ariz.

Micron already stands among the top five DRAM manufacturers in the industry. Collectively, Micron, Samsung, Hyundai Electronics, Elpida and Germany's Infineon (NYSE: IFX - news) control about 80% of the DRAM market.

But troubles in Korea may have a significant impact at Micron's Nampa, Idaho, headquarters. Rumors have been circulating that the two South Korean DRAM giants, Samsung and troubled Hyundai Electronics, may merge.

Hyundai is struggling to reduce some $6 billion in debt stemming in part from the bailout loans it received during the 1997 Asian financial crisis. When Hyundai revealed a staggering $1.97 billion loss for 2000, rumors spread that it might end up being taken over by Samsung at the behest of the Korean government. Such a combination would control 40% of the DRAM market.

That might force other companies, especially smaller Japanese DRAM makers like Toshiba and Fujitsu, to either bow out of the business entirely, or forge new joint ventures like Elpida. But Micron is more likely to stand pat. In 1998 it paid $800 million to acquire the memory-manufacturing assets of Texas Instruments (NYSE: TXN - news), which effectively doubled its capacity, and just last week spent $25 million to acquire KMT Semiconductor, a smaller Japanese chip firm in which it had held a stake since 1998.

But with control of about 20% of the DRAM market, and a knack for manufacturing DRAM chips at very little cost, Micron has little to fear from new DRAM dragons in the East, and little need to make any major acquisitions, analysts say. Micron did not comment for this story.

``If it ends up taking capacity out of the market, it benefits Micron,'' says analyst Jack Geraghty with Gerard Klauer Mattison in New York, who rates Micron a ``buy'' despite a share price that has fallen more than 60% from its 52-week high. ``Would they want to go out and get somebody else? Probably not. For them 20% of the market is enough.''

``This year is looking glum in DRAM, there's no doubt about it,'' says Jamie Stitt, director of business development of Toshiba America Electronics, Toshiba's U.S. unit. He says Toshiba has no plans to alter course by either exiting the DRAM business or launching a joint venture. Instead the company plans to ride out the storm, with hopes of an improvement next year. ``The name of the game this year is survival.''

If industry history is any judge, it usually takes three bad quarters before good news begins to supplant bad, and the DRAM industry is about halfway through the cycle now, says Matas of IC Insights. ``The farther the industry falls now, the better the chance for a rebound next year.''



To: Mathemagician who wrote (67206)3/6/2001 4:36:18 PM
From: Bilow  Read Replies (2) | Respond to of 93625
 
Hi mathemagician; My predictions have been that RIMMs would be cheaper than they are, not more expensive. Maybe you haven't been reading my posts for very long, I've been posting on this thread for years. Maybe I should repeate myself more often, here's what I've been saying about RIMM pricing over the last 12 months. If you think I've been saying that the little darlings are going to remain horribly expensive then show me the (non existant) link:

April 9, 2000
My guess is that eventually the cost of RIMMs will stabilize at somewhere between 1.5 and 2 times the cost of DIMMs. [i.e. 50 to 100% higher than SDRAM] This is in keeping with other industry estimates for the ratio. Right now RIMMs are way out of line, but I don't think that the ratio will stay that way forever. A year ago I thought that the above pricing would have come about long before now, at least by the end of last year. It didn't, and that is something that I found surprising. The high RIMM prices are forcing the development of DDR machines forward a lot faster than they would have otherwise gone on. But even if RIMM pricing had come down to the above ratios, DDR would still take RDRAM out based on cost performance. #reply-13382489

May 5, 2000
The other thing to remember is that the memory makers have repeatedly stated that the cost adder is around 40 to 50%. This is a huge price increase, compared to DDR. #reply-13591290

May 24, 2000
In fact, I expect to see them drop to under 2.0x DIMM pricing, less if there is a big memory shortage as is commonly predicted. #reply-13767231

Sep 25, 2000
(2) Real RDRAM pricing is far below the prices reported in smsperling. (And SDRAM pricing is somewhat below the pricing he reports.) Rambus longs have stated for months that Dell buys RDRAM at a price considerably below the spot price, and this is undoubtedly correct. If the memory makers did start producing for the spot market, they would quickly drive spot market prices to around 75% over PC133 SDRAM prices, I would guess, and I've been expecting RIMM prices to drop to something like this for months. #reply-14451731

Oct 24, 2000
My expectation is that we will eventually see a RIMM penalty of around 50 to 100%. #reply-14645958

Nov 15, 2000
I expected RDRAM to come down in price to around a 50 to 100% adder above SDRAM, but not get below this. It hasn't got there yet, and doesn't look like it's going to ever get there. Maybe it will have a chance when SDRAM goes back up in price eventually. #reply-14812058

Dec 19, 2000
Heck, I've repeatedly stated on this thread that I expect the RIMM premium to eventually stabilize at 50 to 100% above SDRAM levels. That it is still way, way, way above those is something that I find surprising. On the other hand, the price of RDRAM is never going to stabilize at a price lower than that of the standard mainstream memory (now SDRAM, soon to be DDR SDRAM, after that, DDR-II) because RDRAM will always be a niche player. #reply-15049728

Mar 2, 2001
I predicted quite some time ago that RDRAM would eventually drop to about a 50% premium over SDRAM, maybe that will happen someday, but not soon... #reply-15441862

When I make the comments "Wow! The percentage increased! Look out!", it is because I was surprised, I really have been expecting RIMMs to drop to around 50 to 100% over SDRAM for most of a year, not because I am trying to make the local moron's feel bad about their RMBS "investment". Actually, I couldn't give a tinker's dam about their feelings, (as would be clear to anyone reading my commentary on this thread). But in the interim, while RIMMs have been (and still are) incredibly expensive, this has been bad news for RDRAM, and this is slowly decreasing in badness. But the fact is that as long as RIMMs are 1% more expensive than SDRAM, they are not going to become the mainstream memory that Rambus has been promising for so long.

Yes a decrease is good, but this is what has been expected, it is not some incredible thing. It would be incredible if the adder got below 50%, but until then it just ain't news. The Rambus morons think that any kind of a drop is an indication that their memory is the next mainstream standard, but their predictions have been even less accurate then mine, and note that both of us are predicting decreases in RIMM pricing. The fact is that RDRAM is still a (relatively) rare and expensive niche memory. My prediction is that it will remain a rare and expensive niche memory. (Translation of "niche" for mom and pop: RDRAM will remain at around 50 to 100% more expensive than the mainstream memory, in terms of dollars per bit, and won't achieve volumes more than around 10% of total memory production as measured in bits.)

Why don't you make us a prediction as to what the retail cost adder for RDRAM over SDRAM and DDR will be six months from now? Maybe that would be a useful addition to the thread. Predictions as to market share are hard to verify.

-- Carl

P.S. The problem with predicting market share and volume is that no one ever really knows how much of each type of memory got produced in the last quarter or whenever, so you can never get a clear result. The memory makers constantly exaggerate their future production of niche memories. (They do this so as to convince the less intelligent design engineers that the price and availability of the stuff is going to be better than it actually will be.) The only really decent undeniable measurement is spot retail pricing, so that's what I've been reporting. This is in comparison to the other posters on this thread who ignore the subject completely, except for silly generalities.



To: Mathemagician who wrote (67206)3/7/2001 11:23:33 AM
From: Ali Chen  Respond to of 93625
 
"If an increase in the percentage is bad for RMBS, isn't a decrease good? Why not say that rather than shifting the focus to something more convenient to your viewpoint? You do claim to be fair and objective, after all."

Reading your other posts, it does not seem that you are
too dumb. Therefore I cannot imagine how it is
possible to miss the main point of Bilow's price
comarisions: 300% is outrageously high for a five-year
old "technology", especially if it offers no
benefits to PC industry in any area.