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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?


To: Tony Viola who wrote (121974)12/10/2000 10:15:33 PM
From: Ibexx  Respond to of 186894
 
Tony and Intel investors: Intel to test longevity of Moore's Law
By Michael Kanellos
Staff Writer, CNET News.com
December 10, 2000, 5:30 p.m. PT

With a research paper and images from an electron microscope, Intel plans to show that Moore's Law will hold for another decade.

The Santa Clara, Calif.-based chip giant will release a paper Monday at the IEEE International Electron Devices conference in San Francisco showing that the company can produce transistors with elements as narrow as eight angstroms, or three atoms, far smaller than today's chips.

The achievement is significant in that it shows there are no known physical barriers to slow the pace of chip performance growth for the next five to 10 years, said Gerald Marcyk, director of Intel's Components Research Lab.

Moore's Law, formulated by Intel co-founder Gordon Moore, states that the number of transistors a chip can hold will double every 18 to 24 months, as transistor size shrinks. More transistors, which switch off and on to represent binary data, lead to a corresponding leap in performance.

"The trend line is continuing. We are not hitting a brick wall," Marcyk said. "One of the big worries was whether silicon oxides would work at this level."

Chips made with these transistors will contain roughly 400 million transistors and run at 10 GHz, he said. With these chips, computers will be able to translate verbal commands or conversations from one language to another in real time, or search massive and complex optical databases.

These chips will also run on less than a volt of power, less than most of today's notebook chips. Boot-up time, no doubt, will also improve. Current Pentium 4 chips run at 1.5 GHz and contain 42 million transistors.

The future of Moore's Law has massive implications for the industry. For more than 30 years, semiconductor manufacturers have managed to steadily shrink the size of transistors, which has led to chips that are smaller and more powerful than their predecessors. Smaller chips also consume less energy and cost less.

In turn, this has opened the door for software developers and computer designers to create more elegant, powerful products.

Waiting for the law to break
Although the rule has held steady, researchers have speculated about when the laws of physics might stop it. Early in the last decade, Moore himself said the industry probably would hit a wall when transistors shrunk to around 0.25 microns, Marcyk said.

Chips with transistors that size came out in 1997 and are now old-fashioned. Most companies manufacture chips with 0.18-micron elements and will start to make 0.13-micron chips in late 2001.

Transistors of the size being discussed at the conference will show up in processors in 2005, when the 0.07-micron generation arrives, Marcyk said. The average element in these chips will measure 0.07 microns, or 70 nanometers, but some elements will be smaller.

The transistor gate, an open space inside the transistor that controls the flow of electrical current inside the transistor, will measure 30 nanometers wide. The gate walls will be three atoms thick, consisting of "an oxygen, a silicon and an oxygen atom," Marcyk said.

For now, Intel has only managed to produce 0.07-micron generation transistors, not complete chips.

Besides shrinking, chip architecture will change in other ways. The transistors on the 0.07-micron generation of chips will be chemically printed with Extreme Ultra Violet lithography, rather than with today's visible light lithography.

Tantalum or some other element also could replace silicon for the material in the transistor gates, although not the wafer. With a switch in materials, transistors will get wider, but they will leak less electricity, said Marcyk, making them more efficient. Transistors may also be stacked vertically rather than horizontally.

"You can get more of them in a certain area," he said. Still, "a lot of these options are difficult to integrate."

Convention highlights
Intel will be among the many semiconductor companies and academics presenting papers at the four-day convention, which serves to showcase breakthroughs in electrical engineering and present awards to select researchers.

yahoo.cnet.com

Ibexx



To: Tony Viola who wrote (121974)12/10/2000 10:28:04 PM
From: Dan3  Respond to of 186894
 
Re: how can a 20% player dictate ASPs to an 80% player

It can't. But Intel, on its own, can decide to lower prices below what it would otherwise do in order to try to deprive AMD of revenue. Trouble is, as the volume leader, every $1 dollar revenue decrease at AMD is gained at the cost of $4 dollars at Intel. But Intel can afford it. They could just leave their plans as they were prior to the advent of the Athlon, and allow AMD to have ASPs equal to Intel's, but they've chosen to deny AMD market share - and with overall markets not doing well, it may be their best move.

To a degree, both companies have "osborned" themselves. Intel's customers are waiting for SDRAM or DDR P4s, and AMD customers are waiting for 266DDR Athlons. AMD should be out of that hole early next quarter while Intel will take quite a bit longer.

The other reason AMD affects Intel is that the rate at which the actual value of Intel plant is depreciated is increased. When AMD opened Dresden, the value of 4 of Intel's .25 FABs was cut drastically. If not for Dresden, those FABs would have remained state of the art revenue generators for several more years. Instead, Intel has been forced to invest an extra $6 Billion in its FABs to accelerate the openings of planned FABs.

I'm not sure that I'm stating this very well, but IMHO what AMD did was to jump the competitive environment faced by Intel forward by at least a year, with the result that the trailing year's worth of capacity suddenly became worthless (for CPUs).

If IDC is right, and PC growth will be back to near 20% next year, both companies should do well. But in that case, Intel ASPs would fall less quickly, while AMD ASPs would rise substantially.

Just one view, and through AMD colored glasses.

Dan