Bob - Re: "What probability would you assign to Intel conceding this market? "
ZERO per cent.
In 1985, Intel was losing money BIG TIME on DRAM memory. They had been technologically uncompetitive for many years (see note below) and Japan, Inc. was highly focused, highly financed and was eating Intel's - and every American Semi Company's - lunch.
Today, things are way different in the "low end" of CPUs:
1. Intel still has the majority of the sub $1000 PC processor business.
2. Fact 1 is ignored by the media and analysts (and many SI posters) but do the math - AMD and NSM/Cyrix are barely selling 2 million CPUs/quarter combined.
3. Intel is the ONLY x86 supplier THAT MAKES A PROFIT IN THE SUB $1000 PC arena.
4. Intel's technology advances will permit future competitiveness in the sub $1000 area - 0.18 micron process development is nearing completion and 0.18 micron Pentium II devices are being designed - CopperMine.
So, abandoning a profitable business to allow a competitor to make more money is a dumb business decision.
Intel will stay in the market, make money in that market (at lower margins), and insure that NSM/Cyrix and AMD make little or no money - and have difficulty investing in future technology.
Note on DRAMS - Intel had achieved the first functional CMOS 1 Meg DRAMs in 1985 - well ahead of the competition. That device could have been the first competitive DRAM Intel had since the 16k DRAM days in the late 1970s. However, Intel abandoned DRAMS and the CMOS DRAM process development team, based in Oregon, became the new CPU process development team, replacing and surpassing another group in Livermore, CA.
Paul |
