Re: Intel's 23% revenue growth
It's not bad, but it's still trailing the industry:
---------------
SAN JOSE, Calif.--(BUSINESS WIRE)--Aug. 3, 2000--Worldwide sales of semiconductors reached an all time record high of $16.6 billion in June 2000, increasing from $11.2 billion in June 1999, a 48.1% year-to-year increase, the Semiconductor Industry Association (SIA) reported today.
"June's numbers exceeded our expectations and affirm our forecast for a strong 2000," said George Scalise, president of the SIA. "These record-high numbers reflect the growth in the Internet infrastructure and wireless communications markets."
Semiconductor product sectors that experienced significant year-to-year growth in June 2000 included Flash with 167% growth, Field Programmable Logic Devices (FPLDs) with 106% growth, Digital Signal Processors (DSP) at 51% growth, and Standard Linear (Analog) and Optoelectronics at 70% and 65% growth, respectively.
"The consumer applications that these semiconductors are found in include wireless telephones, personal digital assistants, and set top boxes," said Scalise. "The semiconductor industry is critical to the advancement of these markets."
In addition to semiconductor products used to fuel the communications revolution, the DRAM and Microprocessor markets, commonly found in personal computers, also exhibited significant growth from last year. In June 2000, the DRAM market grew 75%, while the microprocessor market grew 36%.
Semiconductor sales were strong in all geographical regions. The Asia Pacific and Japan markets grew 52.8% and 50.8%, respectively, from last year. The Americas market was up 42.7% from last year, and Europe's sales grew 48.1%.
The SIA's Global Sales Report (GSR) is a three-month moving average of sales activity. The GSR is tabulated by the World Semiconductor Trade Statistics (WSTS) organization, which represent some 70 companies.
--------------
I know that microprocessor number most likely includes things such as CPUs for set-top boxes, servers, and workstations, but I'm sure that the primary driver number does come from the PC industry. I think, for the time being, it's best if I don't get into the gorilla vs. king arguments, as it's bound to lead to disagreements, and since I'm not as well-versed on those subjects as many of those here, and may thus create some general misunderstandings. All that I'd like to say is that for Intel's market share bleeding to stop, regardless of whether it's a gorilla, a king, or something in between, one of three things will have to happen:
1. The Williamette proves superior to the Thunderbird solely based on its specs (possible, no way of knowing, however).
2. The world switches to RDRAM to the point where no one's willing to make DDR chips, and the Williamette, being better-optimized for RDRAM than the Thunderbird, performs better (fairly doubtful).
3. (Further down the line) Microsoft refuses to support AMD's x86-64 architecture, leaving AMD up the proverbial creek without a paddle (possible, although I doubt it).
4. Both RDRAM and DDR are produced in large quantities, but RDRAM performs better, and this allows the Williamette to dominate the microprocessor market.
Now for everything that could take place to derail Intel:
1. The Thunderbird outperforms the Williamette regardless of the type of DRAM that's used (possible, by no means a sure thing).
2. DDR takes over the world, and the DDR-optimized Thunderbird profits significantly as a result (very unlikely).
3. DDR performs better than RDRAM, allowing the Thunderbird to take market share (possible, but I doubt it).
4. (Further down the line) Microsoft supports the x86-64 architecture, and AMD, with the benefit of having a backwards-compatible architecture, takes market share as a result.
As you can see, this is a very complicated situation, and you can't analyze it simply by means of a couple of sentences regarding Intel's gorilla power. Personally, I think that RDRAM's going to win out over DDR on performance grounds, but that the Thunderbird will nonetheless remain highly competitive. I also think that not only will Microsoft support the x84-64 architecture, but that Intel might be making a mistake by creating a non-backwards-compatible architecture, and expecting the software industry to follow along without a glitch. While individual companies might make daring moves from time to time, industries as a whole tend to take the evolutionary path of least resistance.
Eric
PS - Here's a few comments from someone at TMF questioning Intel's gorilla power. I don't agree with all of the comments made, especially about RDRAM, but it does make for good reading nonetheless:
boards.fool.com |
