The Inquirer has an AMD roadmap of pricing through Q2 2005.
It’s pretty much in accord with our earlier projections on pricing through 2004.
After that, prices will pretty much go into stasis for the next six months. There will be a little slippage on the lower end of a couple dozen dollars, but that’s it.
There’s two reason why AMD thinks they can do this.
First, their competition isn’t going to be in much shape to do much of anything price-wise in 2005. Intel will be lucky to get a 4GHz Prescott out by the end of the year or early next.
After that, their only move will be to dual-core, which probably won’t occur at all until we’re well into 2005, and such beasts are hardly likely to reach mainstream (i.e. around $200) for the remainder of the year.
This is why we’ve managed to curb our enthusiasm for dual-core so far. We don’t think dual-core will become mainstream technology (and mainstream-priced) until 65nm manufacturing is well underway, which is probably the latter half of 2006 for Intel, later than that for AMD.
This is no great loss given that
a tremendous amount of work needs to be done on the software end before dual-core becomes significantly better than dual processors are today for the average user. So for 2005, any dual-cores that come out will be priced like EEs and FXs, and will hardly making CPUs further down the food chain cheaper.
Second, the PC industry expects demand to be particularly high in 2004 and 2005 because they expect an especially high percentage of older computers currently in use to be replaced (i.e., computers bought prior to the Y2K scare/technology boom). Those replacing such computers are the equivalent of Rip Van Winkles coming back to the market, and the PC firms at least hope they won’t be as finicky about pricing and progress as those more in the loop.
So if you have cows that need to be milked, milk them.
Competition? What Competition?
One of the biggest canards uttered around here is something along the lines of, “We must (love/support/defend) AMD because if it weren’t for them, Intel would be charging us (astronomical amount) for a (very old processor).”
Those days are over. These days, AMD might as well be a division of Intel when it comes to price competition. If that’s what made AMD your “friend,” it’s not your friend anymore.
The old AMD was willing to sell at a discount to Intel to get volume, to give up on price to get sales. The new one will not, and will give up sales to get price.
For Hammers, AMD has kept up price discipline. Since their introduction, the retail prices on Hammers have remained in lockstep with Intel equivalents, basically a few percent off the official retail price.
On the other hand, AMD has not been able to do this with Athlon XPs. They continue to selling at a significant discount to Intel PIVs (but rough price parity to Celerons).
However, this combo has led to AMD’s production mix currently being 85% cheap XPs and 15% expensive Hammers.
In the months and years ahead, AMD’s greatest fear ought to be that, name changes to Sempron notwithstanding, that ratio isn’t going to change very much.
For this audience, I really think that if it boils down to “buy a $250 or more socket 939 chip or a $90 socket 754 Sempron,” I think the vast majority of those reading this will buy the latter.
That’s not what AMD wants, but that’s what I think they’re going to end up with. Rather than shifting their customers upmarket, they’ll shift them the other way, and AMD will end up being in the same old price trap (albeit with a cherry on top).
If The Well Runs Dry . . .
While the PC industry did well last quarter, other electronics firms reported far more shaky results. Combine that with likely increases in interest rates over the next few years, and the computer replacement bonus anticipated may be less bountiful than expected.
This certainly will hurt Intel, and it’s hard to predict what they’ll do about it, but historically, the fortunes of AMD have swung more radically than Intel as the overall market peaks and dips. AMD tends to do better than Intel when times are good, they tend to do worse when times are bad.
Will The Dam Hold?
We know what AMD wants to do, and that is to keep “regular” Hammers fairly highly priced and socket 939 Hammer particularly highly priced.
Can AMD actually do it? That is the question.
If you’re looking for a relatively cheap “real” socket 939 Hammer in the next year or so, it’s clear that AMD doesn’t want to give it to you.
It’s not so clear whether or not AMD will end up having to give it to you.
This article has pointed out some of the reasons why AMD may end up changing their policy. If I had to bet today, I would bet that they will.
However, this is a close call, and I could very well be wrong on this. It most definitely isn’t a sure thing, and I suspect it’s going to take a long time and matters are going to have to get pretty dire before the AMD execs give up on this.