RDRAM penetration, (percentage of DRAM megabyte shipments)is 2% in '99; 22% in '00; 42% in '01; 54% in '02.
To: Carl R. who wrote (49719)
From: Glen2 Friday, Nov 5 1999 9:59AM ET
Respond to Post # 49725 of 49731
From the Dataquest Semiconductor Conference
I've just returned from the Dataquest Semiconductor Conference in Palm Springs. Here is a summary of what was said about DRAM by Ron Bohn, Director of Research, Semiconductor Group:
1. 1999 - Recovery finally started after unheralded decline. This is the first year of the next "boom."
2. 2000 - Demand meets capacity
3. 2001 - Supply constraints. 1995 supply matched. Capital spending rises.
4. 2002 - $63 billion revenue peak.
5. 2003 - Capital additions ramp. The next "bust."
6. DRAM supply/demand tightens in 4Q99 signaling a period of undersupply that will last through 2002.
7. DRAM bit demand will remain healthy over the long term although slower than recent years.
8. DRAM capital spending will be constrained until late 2000 and then rise in 2001 and 2002.
9. By 2003, excess DRAM capacity will be available, causing a revenue decline.
10. Multiple DRAM interfaces will coexist for several years, but RDRAM is the most likely next-generation interface.
11. RDRAM penetration, (percentage of DRAM megabyte shipments)is 2% in '99; 22% in '00; 42% in '01; 54% in '02.
12. SDRAM is 66 MHz for '98; PC100 for '99; PC133 for '00 in low end; faster parts painful.
13. RDRAM will be driven by Intel. A key part of the treadmill.
14. DDRAM will be a niche product for servers, and also used in graphics.
15. Although the conversion from 64M to 256M is beginning, a 128M product will be used by some suppliers because the conversion will be easier and will fill the demand/supply gap in the near term.
16. As is normal in the semiconductor boom/bust cycle, DRAM will see the greatest rates of growth, (compared to other major product sectors) through 2002. Then, DRAM will be the first sector to collapse in 2003. Some things just don't change!
Good trading, Glen |
