Here’s a story which reports some whispering that AMD had a “great third quarter.”
In about ten days, I suppose we’ll get a definition, or more probably, a redefinition of what “great” is.
Nonetheless, there are three good reasons to expect AMD to do better than they’ve done the past eighteen months or so.
One of them is not CPU-related. There was a reorganization of some of AMD’s flash memory operations, which will have the effect of increasing the dollar amounts of AMD business activities and increasing the proportion of AMD business activities in the flash memory area. This could help the bottom line a bit.
The two CPU-related reasons will probably be:
Increased Athlon CPU Sales: There’s been some indirect indicators that AMD is now making rather more processors than they have recently. There are also indications that the PC market will show some decent growth this quarter.
When CPU sales slumped; Intel got a cold, AMD got pneumonia. Its sales dropped much more proportionately than Intel’s. With some degree of recovery, it is reasonable to expect that the reverse will happen and AMD will recover more than Intel.
AMD also got a break with a more-or-less stable pricing structure. Prices of mid- and low-range CPUs stayed stable, so the ASP of Athlons sold in the third quarter probably will stay around the same, too.
A Little Boost From Opteron This will also be the first quarter when Opteron sales will help the bottom line to a significant degree. Sales are probably still miniscule compared to Athlon sales, (almost certainly less than 100,000, likely a lot less), but the ASP and profit margin is far higher, even a relative handful of Opterons sold could push the bottom line in the right direction by tens of millions of dollars.
AMD was only supposed to make about 10,000 FX processors and 80,000 A64s for the quarter, and it’s unclear whether they made that production target, or whether those 90,000 were actually sold by September 30. At best, these sales will have Opteron-like impact, but it will probably be less than that, maybe much less.
It will be next quarter that FX/64 sales should have a big impact of AMD’s bottom line. Production next quarter should be around 300,000 units, and if AMD can hold prices at close to current levels, and can sell their stock at that level, we could be looking at a bottom-line improvement in the neighborhood of $100 million dollars.
If Athlon sales stay healthy next quarter and prices stay stable, all the above could lift AMD into a profitable quarter for the fourth quarter.
Then the opportunity for problems start.
The serious transition to Hammer begins in the first quarter of 2004. At that point, AMD is going to have to cut prices so they can sell the increased number of chips they’ll be making. The first adapters will be history at that point.
The danger for AMD is that they’ll get so hypotized by the high prices they initially get for desktop Hammers that they’ll keep the price skyhigh. This likelihood will increase if Intel feels constrained from planned aggressive price-cutting if Prescott problems continue.
AMD could also get hit by the Apple factor for Hammer chips. The “Apple factor” is a precipitous drop in sales the quarter before a new generation of products are going to be introduced. For probably the first time ever, AMD will be trying to sell a heavy proportion of high-priced CPUs shortly before better comes along. If the second generation of chips come in on time, those aware of what’s coming are likely to wait a few extra months.
That could mean a LOT of unsold chips, and a perception of failure.
Can AMD avoid this? Sure. Outside factors (pickup in CPU sales, Intel’s Prescott problems) make conditions more favorable than could have been expected even three months ago.
What AMD can’t do is get overconfident.