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

To: richard surckla who wrote (59688)	11/1/2000 5:40:10 PM
From: Don Green	Respond to of 93625

Hyundai looking at the end of the DRAM cycle By Will Wade , EE Times Nov 1, 2000 (11:56 AM) URL: eetimes.com SAN JOSE, Calif. — Hyundai Electronics Industries Co. Ltd.'s semiconductor group is reinventing itself as a hybrid DRAM vendor and foundry in an effort to escape the cyclical nature of the memory market and to find a more profitable use for older fabs. Even as Hyundai shifts its business model away from the memory market, however, company executives say they believe the see-saw DRAM cycle is leveling out, as industry consolidation yields a better balance between supply and demand. "We think that with fewer people in the market, and with fabs becoming more and more expensive, the overall market will be stabilized," said Sang Park, president and chief operating officer of Hyundai's semiconductor group. "This could be the end of the semiconductor cycle." Historically, DRAM vendors have made the bulk of their profits during periods of shortage, when they could charge higher prices for products. They then would plow those profits into increasing their manufacturing capacity. When several vendors play that game at once, the proliferation of production sites soon creates an oversupply of DRAM chips, and prices plummet. But Park noted that the DRAM market has consolidated and that only four vendors now collectively account for about three-quarters of all chips sold. That's down from eight vendors in 1996. At the same time, the average cost of a cutting-edge manufacturing site has increased from about $1.5 billion in the late '90s to an expected $2.5 billion for a 0.10-micron fab producing 300-mm wafers. With fewer players committing larger sums of money, Park expects that industry communications and visibility will improve. The long-term result will be a balance between supply and demand. "We believe this will be a very stable market in the years to come," he said. Park's task now is to refocus his company's vision for the long term. That partly entails converting some of the group's older fabs from pure-play memory manufacture to logic foundry work. Successful memory market participation requires state-of-the-art fabs, which today means 0.18 micron and soon will mean 0.15 micron, Park said. While it is possible to convert Hyundai's existing 0.35- and 0.25-micron sites to those advanced nodes, Park's internal cost comparisons show that it would be more economical to buy a small number of tools and convert the older sites to logic manufacture. The company currently derives 5 to 6 percent of its total revenue from logic foundry work and expects that figure to increase to about 10 percent by 2003. "We are recreating Hyundai," he said. "Instead of just a DRAM company, we are a semiconductor company, focused on manufacturing." That may serve Hyundai well in the long term, since not all observers share Park's belief that the DRAM cycle may soon be over. "He seems to think that with fewer players, there's more visibility into capital spending, and the higher costs of fabs will keep out any new players," said memory analyst Jim Handy at Dataquest. "We just don't subscribe to that theory." Handy expects to see memory vendors continue to chase profits during boom times and then overbuild to serve the demand surge, swinging the pendulum back toward overcapacity. "When the profits are there, the memory guys will try to get the most they can," he said. Handy also disputed Park's assertion that the market is more consolidated than ever before. Back in the '70s, he said, there were as few vendors as there are today, and the downturns of the mid '80s and '90s demonstrated that a smaller field does not necessarily lead to smaller swings in supply and demand. "This is not rocket science," said Handy. "The semiconductor cycle will continue."

To: richard surckla who wrote (59688)	11/1/2000 6:15:55 PM
From: Pat Hughes	Respond to of 93625

No Jim doesn't understand. he thinks the 'powers that be' will just dream up a new type of memory and bring it to market. He doesn't realize that if they could have, they would have. Rambus has them all pinned.........and they will all be crying uncle!

To: richard surckla who wrote (59688)	11/1/2000 6:40:00 PM
From: Dan3	Read Replies (4) \| Respond to of 93625

Re: Don't you understand that it doesn't matter what kind of memory Intel uses... I think that you've missed the restructuring the industry is doing that will castrate Rambus and any imitators. First NEC / Hitachi and now Hyundai are separating their foundries from their R&D operations so that if they have to, they can present their demands for their share of IP revenue from any court settlement should Rambus ultimately win. It would be horribly inequitable and economically horribly inefficient if Rambus were granted nearly all of the IP revenues from the industry for next to none of the IP. The existing cross-licensing agreements between combined IP and foundry operations that has been the primary model for the organization of this industry is the only reason Rambus seems to hold an advantage from some perspectives. Were Rambus to win in court, the RAM companies would separate their IP from their manufacturing so that they each would have an IP only operating unit. 3 large RAM manufacturers have already initiated moves in this direction. Cross licensing agreements are contract that can be dissolved by mutual agreement at a moments notice. If Micron, Intel, IBM, NEC, Hyundai, etc. want to dissolve their current cross-licensing agreements and replace them with contracts that require royalties for the use if each companies IP they can do so and it would be none of Rambus's business - other than devastating affect it would have on Rambus's business. It would be painful and expensive for all of them to do this, but if they have to, they will. Dan