Official Conroe information thread

[Sandog]

Originally Posted by specmike

One thing I noticed over the past 2 years or so, give or take, is that AMD seemed to be more than happy to start charing $1000+ for it's flagship FX CPUs, just like Intel was doing with the EE CPUs. However, that may be short lived if the FX series takes a back seat to Conroe.

Mike.

That will have to change if a $400 Conroe beats it. By the time the FX-62 arrives it will probably take the 2.67MHz Conroe to beat it but it should do it well. That chip will probably be in the $600 range at NE(Conroe 2.67GHZ). If Conroe beats AMD's whole line handily there will still be AMD fans who stick with AMD solely because they believe AMD's architecture with the onboard memory contoller and HyperTransport bus is superior to Intel's ancient FSB architecture. It may well be but that just shows how much better the rest of Intels design is doing if it wins in all the benchmarks. The end justifies the means and the end is near. July/August to be exact. It also means when Intel does get around to changing their FSB architecture that we will be in for that much greater performance.

PS. I think the current news has already put the kebosh on people paying $1K for any chips. AMD is already feeling the pain.

[specmike]

Originally Posted by Sandog

That will have to change if a $400 Conroe beats it. By the time the FX-62 arrives it will probably take the 2.67MHz Conroe to beat it but it should do it well. That chip will probably be in the $600 range at NE(Conroe 2.67GHZ). ...........

PS. I think the current news has already put the kebosh on people paying $1K for any chips. AMD is already feeling the pain.

These are the points I was getting to. Can you imagine paying $1000 for a CPU (AMD or Intel) and then having a $400 CPU take you to the cleaners?

I realize how much speculation there is in this area but, we have not even touched what Conroe "might" do if they scale and OC well. I wonder if these CPUs will be able to hit 300fsb+? We may see some pretty insane 1:1 OCs from this CPU.

Mike .

[Water_abx]

Only 95W for the EE Conroe? Very nice. As far as price, oh well. Like I said, a cheaper EE processor is a pipe dream.

[shaihulud]

fans who stick with AMD solely because they believe AMD's architecture with the onboard memory contoller and HyperTransport bus is superior to Intel's ancient FSB architecture

What is interesting that this is not the very first try at integrating the MMU/MMC, and the FSB is not an "ancient" implementation. The integrated controller does not rise to its full capabilities till it is in a multiprocessing system. This does include HT. Single core and dual core SMP will not have much of an advantage. The forte of the Athlon64, really the Opteron, is multiprocessing in long mode. Let the ignorant be asinine.

This is similar to the those who were easily fooled by the exclusive 16-way association L2 cache as better argument. Not one of them understood about the Athlon's 2-way associative L1, victim buffer, and the reasoning for 16 way associative L2. Once explained how exclusive architectures work, it can be revealed how easily people listen to sales talk and buy into it or think it is better.

These are the points I was getting to. Can you imagine paying $1000 for a CPU (AMD or Intel) and then having a $400 CPU take you to the cleaners?

Read my previous posts, you will see that I basically concur. Even if the performance margins are not better economics will always win the most. I feel there is going to be an interesting price war. In other words, it does not matter whom has the better. The cost difference to the consumer will drive much of the market. If the performance margins are so similar, and the cost is higher, who do you think will turn the tide?

The Sharikou blog is nothing but pure biased ignorance, enough said.

[Intel_User]

Anandtech and other review sites tested 2.667GHz Conroe, not 2.4GHz. The 2.667GHz is supposed to be 65W, Woodcrest being said as 80W, and Conroe EE as 95W.

Intel didn't change prices for EEs because AMD beat them, why should AMD?? Historically, that price change didn't happen. Why else would Intel charge only $530 for 2.667GHz Conroe??

It may happen to be shortest performance lead ever. Though Conroe and its derivatives should keep the integer performance crown, we don't know about FP side. Although its really integer that's more important, FP/SSE is what enthusiasts care about, too bad.

[shaihulud]

Intel didn't change prices for EEs because AMD beat them, why should AMD?? Historically, that price change didn't happen. Why else would Intel charge only $530 for 2.667GHz Conroe??

The key to this is in my previous post, just before yours. The second to last segment.

AMD has not had competitive opposition for what I would say a good year and a half. This allows them to charge a premium. This is simple economics in the working. Now, if the Core is able to be just as or even more competitive and at the price range that Intel is declaring. Then AMD will have to compensate with the opposition. However, I feel the price changes should not effect the Opterons as much.

This will also mean that AMD will not be so lucrative as they are now with the processor market. Lower costs to Intel will not be so detrimental for they can easily soak up far more loss in revenue. This move by all means is nasty. Come back at lower costs, win back market share, gain fanaticism, whatever-this is capitalism cutthroat. This battle will happen over 4 quarters, unless the new architecture can be released sooner by AMD. If the new architecture is not comparable this can backlash on AMD and compound the effect.

I think even the moniker of Core is psychological. Now, every tech site will have to be more descriptive or definite when using the word. Did they mean Intel's Core or just core in general. This will change the vernacular used, and I feel is kind of dirty. Any mention of core now can be disconcerting. The K8L core , the Pentium 4 core, or how about just the Core.

This is going to be an interesting 4 quarters on all fronts. New compilers, chip sets (didn't want core logic to be confusing), motherboards, memory prices, etc. This is kind of invigorating, and hopefully will change the expenditures that I will make soon to not be so high and get so little. No matter what architecture I decide to purchase.

[Sandog]

There are a couple of things related to cost. First Intel is ahead on .65nm. This is important as chips cost roughly half as much as they do at 90nm. This should enable Intel to offer reduced prices. Intel should have the Conroe (very capable Vista chip) in quantity and with a price/performance ratio better than AMD's desktop chips. The other thing is AMD has a bad history of meeting chip demands. The same thing will happen to server chips. Intel will introduce the Woodcrest in quantity at .65nm which should cost lower than the Opterons to produce and perform as well. Intel was said to have gone to 90% .65nm by the end of 2006. This is speculation on my part but this is what I see happening.

[Intel_User]

Kentsfield might be brought as early as December of this year. It means Intel better intro Conroe at mid-2006, not late so Conroe can sell for some time. This 4-month gap in introduction practically the whole 1 year is gonna become pretty straining I bet. I wonder what would happen to Conroe EE. Would they be able to get 3.33GHz Conroe EE's same time as mainstream Conroe's?? Historically such huge gap between EE's and mainstream's haven't been really done.

The key to this is in my previous post, just before yours. The second to last segment.

AMD has not had competitive opposition for what I would say a good year and a half. This allows them to charge a premium. This is simple economics in the working. Now, if the Core is able to be just as or even more competitive and at the price range that Intel is declaring. Then AMD will have to compensate with the opposition. However, I feel the price changes should not effect the Opterons as much.

You are right. But currently, there is no information around that suggest that AMD will change the ASPs.

[shaihulud]

Yes, a lower micron will reduce cost to the manufacturer and therefore reduce price to the consumer. I do not think that it is at such a dramatic half reduction you speculate. My statements and arguments are not at a physical level. Although, they do contribute such as smaller die space. My argument can be understood when both Intel and AMD have had a lower die shift there was not a very high reduction to the price for the consumer. If one existed it was not but usually in the 10's to 20's of dollars. The law of supply and demand contributes more to cost than a smaller implementation of lithography, or whatever method is used.

AMD may not meet the demand, but they are scaling to meet the demands. This is a good thing IMO. Demand is high and supply is low. Therefore a premium can be extracted. Since there is not a competitive front, and with the economical factors the premium will exist till there is better competitive option. If the next generation of architecture proves to be competitive. Then AMD's supply will not hurt them as much as lower revenue for demand and supply will be low.

The reason why I think the Opteron will not be effected as much, and I can be absolutely incorrect, is due to the fact how well Opterons scale when configured for multiprocessor systems. I am unsure how Intel's new architecture will work in multiprocessing. Contention prevents scaling, but the Opteron integral features and I/O buses help alleviate this issue. Beyond SMP I am unsure how well it [Intel] will exactly perform, and am very curious. Any input on this?

You are right. But currently, there is no information around that suggest that AMD will change the ASPs.

We will not see this till supply and demand take over-economics.

[Sandog]

I was always careful never to say half which I never did, I have heard talk before of what the difference was but of course it is not half. Mathematically, there is a 38% reduction in size between the two dies. Intel also went from 12" to 16" wafers I believe which is a 33% increase in diameter. Intel also has a lot of fabs running. The cost per chip is substantially decreased for Intel. How much savings Intel is willing to pass on to the consumer is up to them. Typically you sell for the market value and hope it covers costs. I believe Intel will undercut AMD's chip prices and still have a big profit margin. They will continue to do this until AMD can get their .65nm parts available and meet demand which they never seem to do. AMD is busy with .65nm also and may even quadruple total chip production by early next year but what is the demand for AMD going to look like by then??? All in all I think Intel will be able to stem the flow of blood. The market shares Intel lost in the recent past will probably never be recuperated. Like another poster said, it won't come down to performance much anymore as both companies are building great chips. Market share will come down to which company has the best value for the dollar. Intel has a reputation in the industry for quality and they also have better support IMO. This should play a major role in buying decisions since they add a lot of value to the product. Intel is also more adept at marketing their products. To me Intel just plain rocks any which way I look at it.

[h20-ski]

I think that buy the time AMD gets 65nm figured out Intel will be starting production on 45nm chips making more of a good thing.

[shaihulud]

Intel is ahead on .65nm. This is important as chips cost roughly half as much as they do at 90nm

Was I not in context?


I do agree that the over all package that Intel can deliver is better. In the many testings that I have done I found that the Intel products are better, IMO. I should note that many criticize AMD for not being similar, and companies like VIA have tainted AMD with horrible core logics.

Personally, I like the fact that I can read whitepapers on anything that is made by Intel, and I should note that AMD documents well too. I dislike the fact that companies such as nVidia have not released such information concerning their products.

I always disliked the naming schemes used by AMD. 3DNOW! is not exemplary, and is horrendous. MMX or SSE is excellent and succinct. Thankfully x86-64 was perfect, and I understand the moniker change to AMD64, which is not bad, once EM64T came around. At least AMD64 was not 64Today!

Market share will come down to which company has the best value for the dollar

It always does....

[Sandog]

Originally Posted by shaihulud

Was I not in context?

...roughly

All this speculating is making me nauseous...must have hardware soon...lol.

[Sandog]

I really think the best time to upgrade will be after Vista comes out. Everything is going to change with that operating system, software and hardware. Intels 965 chipsets were designed for Conroe and Vista but until the OS is here and you can run it fully you will never know. Especially when MS is changing 60% of the code. Intel probably started the 965 chipset mid 2005. About the only safe hardware to buy 'till then is a hard drive lol.

[Sandog]

Originally Posted by shaihulud

I do agree that the over all package that Intel can deliver is better. In the many testings that I have done I found that the Intel products are better, IMO. I should note that many criticize AMD for not being similar, and companies like VIA have tainted AMD with horrible core logics.

Personally, I like the fact that I can read whitepapers on anything that is made by Intel, and I should note that AMD documents well too. I dislike the fact that companies such as nVidia have not released such information concerning their products.

I always disliked the naming schemes used by AMD. 3DNOW! is not exemplary, and is horrendous. MMX or SSE is excellent and succinct. Thankfully x86-64 was perfect, and I understand the moniker change to AMD64, which is not bad, once EMT64 came around. At least AMD64 was not 64Today!

It always does....

Intel provides design guides with their chipsets that explain how to lay everything out. What a gimme. If every manufacturer follows the circuit layout that Intel provides they would have a board similar to an Intel board and the least problems. I know ASUS does a good job. It's called support.