Originally Posted by melgross
You're repeating the same assumptions, that you can't support.
You are making up some numbers, and using others that aren't relevant. (1)
As I mentioned earlier, you can;t use numbers from the end of a chip line to numbers at the beginning of a chip line. The old chips have been dropped in price several times, while the new ones are at the highest they will ever be. (2)
Also, you are forgetting that Apple rarely lowers the price of its models once released, unless they make major changes during that time. (3)
It's very dangerous to make assumptions based of numbers that may not have anything to do with Apple's prices. You don't know how much Apple pays for these chips. Prices released are for bin quantities. Apple, and other large manufacturers, buy many bins of these chips, bringing the prices down a long way.
It's very possible that Apple's price for these chips is no more than half the numbers you quote, and therefor, make up a much smaller percentage of the machines cost than you assume. (4)
Perhaps you didn't notice it, but the last column in which the prices are given are clearly headed "Release Prices". They don't reflect current pricing. (5)
The only correct thing here is the part about the 2.4 GHz part. (6)
You are under the impression that Apple WILL release an i7 Mac Pro? They won't. (7)
There is NO Xeon labeled i7. The i7 is purely a desktop chip designation. Xeon is a server and workstation designation. The two are not interchangeable. (8)
95%? As far as you understand? Where did you get that understanding from? A guess? I've never seen any analysis of Apple's desktop sales that have come anywhere close to that number. 60 to 75% possibly. (9)
According to your "understanding" I'm surprised you aren't simply calling for its discontinuance. (10)
(1) I think you're the one making most of the assumptions, without any proof or irrelevant links than don't concern the Mac Pro Xeons. See below.
The only numbers I may have made up are:
- the price of the server chipset ($100): this is based on the knowledge of the price of mobile/desktop chipsets that can be found easily and that are priced between $20 and $70. A server chipset is of course more expensive, but doesn't compare to the price of 2 high-end server cpus which Apple uses on the Mac Pro.
- the % of Mac Pro+Mac mini va iMac sales. See below.
(2) End of the line or not, if you took the time to look at Intel's price list, you would have seen that the list price of the Xeon cpus for the Mac Pro has not decreased. In fact, most of the DP-Xeon cpus have not seen their list price change since their launch (woodcrest included). So, unless you have actual documents that prove your point, I trust Intel's price list over your assumption.
(3) Of course not I'm not forgetting. The fact is that since the Intel switch, Apple keeps increasing the entry price points on all their models: Mac mini was $499 now $599 (and old), MB was $999 then $1090 now $1299, iMac was $999 then $1199, Mac Pro $2199/2499 then $2299/2799. I'm the one thinking that Apple will increase again the entry price point of the Mac Pro to (probably) +$3K for a dual 2.66 models with 6GB RAM. Now if Apple chooses another cpu or a different RAM configuration, price may differ.
(4) Of course Apple doesn't pay the list price. But I think you shouldn't apply your general guess of "no more than half" on the Mac Pro. Apple is not selling millions of them, so they are not buying millions of xeons. The discount may not be as much as you think. But no matter what the discount is, if the list price for a chip "A" is higher than the one of a chip "B", if Apple buys the same quantities, they will pay more for the chip "A" than for the chip "B".
(5) That's exactly why i linked those list. They show that the "Release Prices" and the current prices are the SAME for the xeons used in the Mac Pro. If you had bothered to open Intel's list price, you would have seen that.
(6) The only correct thing here is... well you couldn't even get this right.
(7) No I'm not, I think a Core i7/W3500 Mac pro would be a nice addition to the Mac Pro family. I've never said Apple will make a Core i7 Mac Pro, I think they should.
(8) You've got it the other way around. That shows how much you are paying attention. The current X3300 series of Xeon chips are in fact desktop Q9000 series cpus with a Xeon label. They have the SAME SPECS, the SAME PRICES and the SAME PRICE CUTS. Go to the Intel site and see for yourself. The future W3500 series of Xeon chips will be in fact the desktop Core i7 cpus with a Xeon label. They have the SAME SPECS and the SAME PRICES.
(9) Since Apple doesn't disclose model numbers, your GUESS is as good/bad as mine. I'll tell you that my "guess" was based on unofficial numbers for the year 2007 that gave 3% for the Mac mini and 8% for the Mac Pro. My "assumption" for the last 2 quarters is that those numbers were even worse, hence 90-95% (not just 95%) for the iMac. Even Tim Cook said that the Mac Pro was not a significant part of the desktop business. I'd be happy if you could share actual numbers...
By the way, since english is not my mother tongue, I may sometimes use expressions in an odd way...
(10) No I'm not. I think that there is a lot more to do in the desktop business, and for me, the numbers from last quarter just show how bad things can get when you don't take care of it (the desktop business). What I would like is for Apple to release new/updated models as soon as possible.