RE: But I think AMD is a far better off than Intel, mainly because they made their big investment in .13/copper/Pod capable manufacturing plant 2 years ago and are now entering the cash cow phase for that plant. Intel, OTOH, has to make a huge investment in plant to stay competitive with AMD, and make it a time when nobody's going to see much return on any semi plant investment.
---First off, AMD's "investment" was 0.18u, not 0.13, and "pod" capability has been built into every lead intel Fab since 0.25u was in existence. Intel is already in 0.13u manufacturing production in Oregon, with Fab22 most likely coming on-line in early Q1/02, right on time for when most Analysts agree should be a turnaround.
There is considerable new capacity coming on line in Taiwan, Korea, etc. just as utilization rates for existing capacity have dropped to 50%. Intel has usually been able to produce marginally profitable parts from its one generation old FABs, and this has let them be successful with their policy of adding rather than replacing foundries. Not this time
---Hardly.... Most of that Asian Capacity is in commodity chips, and chipsets, which intel can counter with their remaining 0.25u Fabs, and later on, when P859 gets rolling, the 0.18u Fabs as well. Most, if not all of those Fabs have been paid for many times over, while "Asia" is facing major debt for building their Fabs.
Intel shot its wad last fall/this spring with P4, their first major new X86 architecture in 6 years - there's another long dry spell ahead. And they have to practically give the things away. Merced/Itanium/McKinley is two years late and trying to launch into the worst market for servers in recent memory. Merced faces a number of proven competitors, each with a substantial installed software base. This is at a time when those competitors are willing to sell below marginal cost to minimize layoff costs from their manufacturing employees..
---Nowhere to go, but up..... P4's upcoming 0.13u "Northwood" introduction, should fix all of the complaints about the 0.18u P4 version. Whatever money Itanium makes in the marketplace will just be "profit gravy", the penetration again, paving the way for 0.13u McKinley version. By the time AMD gets it's "Yesterday’s Technology" "Wheeze-Hammer" out of the gate, the IA-64 steamroller should have been rolling for quite some time.
That plant is worth plenty, but nowhere near $17 Billion. Intel has to cover the costs of the new plant they're building, and find funds to cover the inevitable write off of 1/3? 2/3's? of that plant, all during what is being described as the worst semiconductor downturn that has ever taken place.
---Intel has reiterated it's commitment to spending 7.5B this year in Capital Expenditures. Of that, ~2.3B is going towards building Fab22 in Chandler/Ocotillo Az, most likely a good chunk is also being used for 300mm Wafer Development, and some to update the 0.13u Technology to it's "2nd Stepping". Again, all well ahead of similar AMD efforts. There won't be a repeat of "Coppermine" shortages this time.
AMD has been seeing decent demand for some of its products recently, and is probably missing the opportunity to increase market share because they don't have enough plant. But AMD is smart enough to know that for the next year, and possibly the next 2 years, the name of the game is keeping costs down - and they are doing that even as, incredibly, they are maintaining or increasing their market position.
---And I think that AMD has been riding the "Athlon" coattails" a bit too far. Besides revisions of that core, and the declining profit margin flash business, what else have they got?? So far, it seems that Athlon SMP is not seeing wide availability OR acceptance, and the whole DDR Memory Hype, has been a bust. In contrast, Intel is on the verge of introducing New Products, and new Process Technology, while instituting strong cost savings programs at it's 0.25u and 0.18u Fabs.
Timing may not be everything in this business, but it is very important. It may have been pure luck, but AMD's timing has been perfect, and Intel's has been perfectly awful.
Regards,
Dan
---You're right..... Timing "IS" everything...... If you REALLY think about, you'll come to realization of the truth. AMD had a BIG opportunity to KO intel, and I got to admit, WE took one to the Jaw, and hit the canvas....
The thing is...., we got back up before the 10 count. Unfortunately, AMD didn't have the follow through KO Punch. We made it to the bell ending the round.....
And Now.... We're REALLY Pissed Off.....
:-)
Semi |
