Intel price cuts

[Idontcare]

The desktop Core 2 Duo E8500 (3.16GHz) got the largest cut, from $266 to $183 (31 percent) on the new pricing list, dated July 20.

The next biggest price reduction was the desktop Core 2 Duo E7200 (2.53GHz), reduced 15 percent, from $133 to $113.

Other price cuts include the popular quad-core Q6600 (2.4GHz). It drops 14 percent, from $224 to $193. The desktop Core 2 Duo E8400 (3.00GHz) falls 11 percent, from $183 to $163.

Xeon server processors also saw price cuts. The X3220 (2.40GHz) was cut 12 percent, from $224 to $198. The X3210 (2.13GHz) saw an identical cut, while the E3110 (3.00GHz) was reduced 11 percent, from $188 to $167.

Intel cuts chip prices up to 31 percent

 

[The-Noid]

Looks good.

 

[BRDiger]

Still doesn´t seem 2 affect the german market as now... E8500 for example stoll goes for the same price as on July 1st...

 

[HOOfan 1]

What about pricecuts on the Q9550?

What about the introduction of the Q9400, the Q9650 and the E8600?

 

[OCGuy]

I dont think the yearly price-cuts neccesarily coincide with the product release slated for Q3, which will slide the price brackets down a notch.....

 

[Denithor]

Yet another round of "hold AMD's head underwater until they drown" or what?

I mean, don't get me wrong, I love paying less for higher performance as much as anyone else here, but jeez, how is AMD supposed to ever become competitive again if they are forced to keep cutting prices just to match Intel's superior chips?

 

[Idontcare]

Originally posted by: Denithor
Yet another round of "hold AMD's head underwater until they drown" or what?

I mean, don't get me wrong, I love paying less for higher performance as much as anyone else here, but jeez, how is AMD supposed to ever become competitive again if they are forced to keep cutting prices just to match Intel's superior chips?

Same way Via and Apple have managed...you find your niche and scale down your company's size so as to match the revenue stream you can expect to extract from your niche.

From a technical point of view, simply given the costs associated with continued process technology node scaling (32nm, 22nm, etc) there really is no practical way AMD can be expected to continue to maintain even the current 1 year behind Intel cadence.

That's assuming they can maintain a design team capable of creating CPU architectures that are IP competitive with Intel.

Add together the combined synergistic effects of Intel fielding 2+ design teams in parallel and those teams having access to a fully funded process technology pipeline and there is just no technical merit to AMD recovering unless Intel just somehow massively majorly blows it (and Otellini just doesn't seem the type to do that).

At least the ATI side of AMD has a chance because NVidia has no greater access to advanced process technology than ATI does (they both use the same foundries). So in this sense it is purely a GPU design race and the barrier to entry in this race is still well within AMD's financial capability (until Larrabee rolls out, if it is all that Intel wants it to be).

But the CPU race seems all but officially over, if for no other reason than Intel will have 32nm out probably for 2yrs before AMD does and if the gap at 32nm isn't 2yrs then it most surely will be when 22nm rolls around. No one can out-design a competitor's design team who has access to superior process tech...Intel showed that at 90nm.

 

[The-Noid]

As you know Idontcare the most important market to AMD and INTC is the enterprise market. AMD needs to find a way to be competitive in the 2P-8P market which is where they were making the majority of their profit, when they were profitable. However, with the new IMC of the new INTC products I question whether that will be something that can be done.

 

[Idontcare]

Originally posted by: Yoxxy
As you know Idontcare the most important market to AMD and INTC is the enterprise market. AMD needs to find a way to be competitive in the 2P-8P market which is where they were making the majority of their profit, when they were profitable.

I agree on all counts. I know the 2P-8P market segment is high(er) margin, those $2k 8xxx opteron SKU's are proof of that, but what I lack knowledge of is the revenue volume of that segment.

Is it a $2B annual revenue segment or a $0.5B or perhaps $10B?

Originally posted by: Yoxxy
However, with the new IMC of the new INTC products I question whether that will be something that can be done.

And hyperthreading...if it is anywhere close to as effective as the hype has led us to believe then things get even more precarious for AMD in the 2P-8P segment when Nehalem comes to town. Intel appears to have left no stone unturned when it comes to Nehalem and the server segment.

 

[magreen]

Originally posted by: Idontcare
No one can out-design a competitor's design team who has access to superior process tech...Intel showed that at 90nm.

I'm not following you here, Idc. What did Intel show at 90nm? That was the p4 vs. A64 era wasn't it, with the A64 well in the lead?

 

[nyker96]

nice, I see newegg has some price adjustments already primarily e8500s, e8400s, q6600s.

 

[Idontcare]

Originally posted by: magreen

Originally posted by: Idontcare
No one can out-design a competitor's design team who has access to superior process tech...Intel showed that at 90nm.

I'm not following you here, Idc. What did Intel show at 90nm? That was the p4 vs. A64 era wasn't it, with the A64 well in the lead?

That's my point precisely. At the time AMD certainly had bright innovative architecture engineers, but they also had access to arguably superior process technology in many critical aspects at the time.

AMD's 130nm with copper interconnect (Intel didn't transition to copper until the 90nm node) and subsequently AMD's 90nm node with SOI (leakage, or lack of it, was the winner for A64 and the loss of Prescott) were superior process techs for their time and place and enabled the K7 and K8 to make the performance wins they did.

Now the shoe is on the other foot (Intel is rolling more advanced process tech now) and the ability of AMD to ever regain the technology advantage like they had circa 130/90nm nodes is just prohibitively expensive when attempted at the 32/22nm nodes.

The term here is "barrier to entry", the barrier to entry for process technology is eclipsing AMD's ability to pay for it (it even eclipses the IBM fab club's ability to maintain the cadence as set by Intel).

 

[Extelleron]

Originally posted by: Yoxxy
As you know Idontcare the most important market to AMD and INTC is the enterprise market. AMD needs to find a way to be competitive in the 2P-8P market which is where they were making the majority of their profit, when they were profitable. However, with the new IMC of the new INTC products I question whether that will be something that can be done.

The 4P+ server market is profitable, and AMD controls the majority of it right now, but it alone cannot sustain a company nor can it out weigh bad margins on the desktop. Selling 5,000 Barcelona's for $2,000 each and selling 500,000 Phenoms for $200 each still give you an ASP of $217. And as for actual Phenom/Barcelona ASP.... that's way higher than reality. Most Phenoms sold are probably X3s or the X4 9550, so desktop Phenom ASP is probably in the $150-170 region. Most Barcelons sold are going to be the 2xxx series, not the high margin 8xxx series. I'm getting into a rant here but the bottom line is that the sale of a few thousand 8xxx Opterons isn't going to matter much in the face of low ASPs on the desktop.

As Idontcare has stated, I believe that the performance race between AMD and Intel is over as far as desktop consumers are concerned. In order to win a company needs either superior process technology or a superior design, neither of which AMD has now or seemingly will have in the near future. The superior process, that I can tell you right off the bat AMD will never have and has (as Idontcare has shown, this is arguable @ 90nm, where Intel suffered. But there was never a problem with 90nm Pentium M CPUs AFAIK....I think Intel's 90nm problems stemmed just as much from Prescott's bad design as they did from a lackluster 90nm process). But right now there is no question Intel is superior... they are currently 1yr ahead on 45nm and a solid 2yrs ahead on hk/mg, and that's assuming AMD converts to 45nm HK/MG in 2009. If they wait till the 32nm node.... that's 3yrs behind. And the gap is only going to get wider as AMD can't afford the R&D anymore. The IBM alliance may help a bit, but the best interests of IBM and the other companies don't always match the best needs of AMD.

As for design.... right now they are not even close to Intel, at least once Nehalem rolls out. Right now you can argue that Phenom is a theoretically more advanced design than Core 2, but the performance is just not there. Even if Bulldozer is amazing, by the time it comes out, Intel will be readying Sandy Bridge.

 

[Idontcare]

Originally posted by: Extelleron
Even if Bulldozer is amazing, by the time it comes out, Intel will be readying Sandy Bridge.

When you put it that way it is downright depressing to think about the prospects of AMD not only having to fight Nehalem with Shanghai but Westmere as well.

 

[Extelleron]

Originally posted by: Idontcare

Originally posted by: Extelleron
Even if Bulldozer is amazing, by the time it comes out, Intel will be readying Sandy Bridge.

When you put it that way it is downright depressing to think about the prospects of AMD not only having to fight Nehalem with Shanghai but Westmere as well.

Given Intel's rapid release schedule with tick-tock, there is just no room for a "slower" company like AMD to catch up. Everytime AMD gets close.... there is something new from Intel to best it.

On the desktop it does look like Deneb/Propus are going to have to duke it out with Nehalem/Westmere, although we don't know much about AMD's plans in this sector. On the server..... it's definitely K10.5 architecture into 2010 with San Paolo/Magny Cours. As exciting as 12 cores on one package sounds, I doubt it will hit the frequencies necessary to best an 8-core Westmere..... or however many cores Westmere ends up having (There is a rumor I heard somewhere that Westmere is 6-core, but who knows).

At this point who even knows what will be going on with AMD in 2010 or 2011, I don't think AMD knows for sure either. I don't know what is going on there but I would assume it is more of a "paycheck-to-paycheck" atmosphere than a forward looking one.

 

[OnePingOnly]

Did the Q6600 already come down in price?

 

[Idontcare]

Originally posted by: Extelleron
On the server..... it's definitely K10.5 architecture into 2010 with San Paolo/Magny Cours. As exciting as 12 cores on one package sounds, I doubt it will hit the frequencies necessary to best an 8-core Westmere

It seems like a crazy tough challenge to get 12 cores of 45nm goodness to fit into a 140W TDP budget at a clockspeed that will be competitive with 32nm Westmere.

Originally posted by: JRussellDMD
Did the Q6600 already come down in price?

Retail Q6600 are $199 at newegg now, not too much of a premium over the 1k tray price.