Abit recently made a few provocative comments in a recent newsletter.
“Certain enthusiast motherboard companies have seen their ratio of AMD-Intel products sold go from 70/30 in favour of AMD to almost the exact opposite.”
Intel has also said that its processor sales are a little better than expected.
Are these indicators that the tide is turning, that enthusiasts are turning from AMD to Intel?
I think not.
What Is Happening
One should never confuse a momentary surge due to an obvious cause for a long-term trend.
Intel just released some whiz-bang products. Those who were waiting for them are now starting to buy them. Of course the number of Intel sales are going up.
On the other hand, at least for this community, AMD did what it was going to do a while back, and people have either already upgraded to TBredBs/nForce2s or are waiting for nForce2 Ultra 400s/KT600s and/or waiting for Bartons to get really cheap. Others are waiting for Athlon64s.
So a lull in AMD sales by enthusiasts is perfectly normal at this point, too.
Combine an Intel surge with an AMD lull, and sure, you can get this kind of turnaround in numbers, but it doesn’t mean anything long-term.
This is not really a zero-sum game. AMD sales may very well decline throughout the year in general and among enthusiasts, but that doesn’t mean they’re buying Intel. On the whole, it means they’re just not buying, period.
When you look at the forums, you just don’t see a ton of “I’ve had it with AMD; I’m going back to blue” comments out there. For that matter, there really isn’t any huge surge in Intel activity, either. There’s interest, there’s a bit of erosions, but no signs of a Big Shift like there was in the TBird Era.
A big reason for that is the audience’s increasing relative satisfaction with what they already have. People no longer want just improvement, they want an improvement that will blow their socks off.
Up to a few days ago, my work machine was a TBredB running at 2.3GHz-ish on an nForce2 board, one stick of PC3000, Radeon 9700 Pro, WD1200JB.
Now I’m running a PIV at 3.5GHz, two sticks of PC3500, two RAIDed Raptors and the 9700.
Do I notice an improvement? Occasionally. Does it blow my socks off? No, it doesn’t even clear my nasal passages. 🙂
Yes, it’s better. Better enough to justify the expenditure for the average person reading this? I don’t think so.
If there’s going to be a Big Shift back to Intel in the foreseeable future; it will only happen if and when AMD falls flat on its face with Athlon64s. Not even the first go-round this fall; the second go-round at 90nm.
In all honesty, even if Athlon64s aren’t hyper-competitive with Prescotts, so long as AMD keeps them cheap and at least somewhat close in performance to Prescotts, there will be no great defection.
In short, Intel can’t take the ball away from AMD. AMD is going to have to drop it, and charging a lot is a far more likely cause of a fatal fumble than a somewhat slower processor.
Horseshoes, Not Raising
In the next couple weeks, I’ll be working with Canterwoods and Springdales to first get some more complete answers to the memory bandwidth issue we raised yesterday, then look to see if Canterwoods are really worth the extra cost over Springdales.
After that, though, what I plan to do is take my old system and compare it to my new system.
No doubt AMDroids will scream bloody murder about bias and unfairness, and it would be if people just considered performance, but real people don’t do that, at least not these days.
People in this audience don’t buy strictly on performance any more. Cash outlay is a big to huge to only factor for most. The buying decision people really make is, “Should I buy a cheap TBredB/nForce system, or spend a lot more on a 2.4C and Springdale/Canterwood? Will it be worth it to me?”
Or, if they already have the first, they want to know if the second is worth it.
These people know the first isn’t as fast as the second, but a couple hundred dollars is more important to them. They might want to know how big the gap between the two is, but so long as the gap isn’t too big, it bothers them not.
So that’s what we’ll do. We’ll do a comparison where we know even before we start that AMD will lose most of the time, simply because in this competition these days, it’s not a footrace. Rather, it’s like horseshoes.