EU conducts antitrust raid on Intel and retailers

[Modelworks]

Originally posted by: Nemesis 1
Couldn't Via if they want start a lawsuite against intel and AMD , After their new chipis out for a year. Saying AMD slashed prices and lost money in so doing . From what I understand Via chip is = to Dothan . Thats not a bad chip at all. But to compete against AMDs prices it is hurting Via. Yes I know Via chip is new . But in a year couldn't VIA make the same case against AMD's present pricng. When Via comes out with their 2 core chip it gets even more interesting.

Give a theif enough rope and they will hang themseves.

Not really.
Via has not been in a position to really be considered a competitor.
It would be like Toyota taking Porsche to court. They are both in the car market, but target different customers.

Via is aimed at the low end, average performance pc market. They aren't trying to be the cpu for gaming or number crunching. They remind me a lot of cyrix. Except cyrix wanted it all. Cyrix has ties to via and amd.

If you want to read a really interesting story of a cpu company that fought Intel hard and made x86 without Intels permission, check out:
http://en.wikipedia.org/wiki/Cyrix

 

[CTho9305]

Originally posted by: Nemesis 1
So many pointing at intel saying they selling at below cost . or rebates. Yet point out to me a year they lost money or even a qt. Seems to me they can sell at any price they want as Long as they make money on the chips.

Did you never read AMD's complaint? One Intel's monopoly tactics looks like this:
1. You're an OEM who can sell 100,000 computers each year.
2. You've bought 90,000 CPUs from Intel at $100 each. AMD isn't big enough to supply all of your demand, but can cover a big chunk of it.
3. You're thinking about who to buy the next 10,000 CPUs from. AMD offers a faster CPU for $50 each, but Intel says, if you buy 100,000 Intel CPUs, they'll retroactively drop the price on every CPU to $89.
Even if AMD's CPUs were free, you'd actually make more money by buying 10,000 more Intel CPUs (100,000 * 89 vs 90,000*100 + 10,000*0).

At first glance, the deal doesn't look that evil - it looks a lot like like an ordinary volume discount. However, when you look again, you can see that it's set up in a way that makes it impossible for competitors to enter the market. A competitor can only get in if they have the capacity to supply almost all of the market (so that OEMs wouldn't have to buy CPUs from Intel at all). A competitor can't break into the market slowly. See the section starting with paragraph 59, on page 22 of the complaint.

Intel didn't slash prices when AMD was on top in performance . Why because AMD couldn't sell anymore chips . They couldn't make them fast enough.

1) If that's true, why did Intel feel the need to use its dirty tactics? Are they just that greedy? Seriously, read the complaint. There's a lot of bad stuff in there.
2) I don't think that's even true. If you look at the AMD complaint, AMD offered Hewlett Packard 1 million CPUs free, and couldn't give them away. If Intel was competing fairly, it would have been a pretty strange decision for HP to choose the more expensive product. See my explanation above as to why anti-competitive behavior by a monopolist could make an OEM reject free CPUs.

Why didn't AMD have Charter make chips befor C2D was released.

That agreement was made back in 2004. Last time I checked, 2004 was before 2006, which would mean it wasn't motivated by Core 2. Somebody might want to check my math though .

 

[Viditor]

Originally posted by: Nemesis 1
So many pointing at intel saying they selling at below cost . or rebates. Yet point out to me a year they lost money or even a qt. Seems to me they can sell at any price they want as Long as they make money on the chips.

It has nothing to do with selling below cost in this case...

1. Intel offered their rebates retroactively based on individual retailers total chip sales. This means (in essence) that at the end of the month, Intel would give money to the retailers if they didn't sell AMD chips.

2. A monopoly isn't illegal or wrong...it's the abuse of a monopoly that's wrong and illegal. In this case, Intel used their monopoly to effect this "marketing strategy". Please note that AMD could not have done the same thing because the retailers could not survive without Intel parts...

 

[zsdersw]

Somehow I doubt the "damages" to AMD are terribly significant.. definitely not worth $4B, let alone $20-30B. Considering AMD was immediately selling every chip they made.. *before* they started getting systems with their chips inside offered by the likes of Dell.. I find it hard to believe their bottom line would've been improved that much if that of which Intel is accused didn't occur.

 

[Viditor]

Originally posted by: zsdersw
Somehow I doubt the "damages" to AMD are terribly significant.. definitely not worth $4B, let alone $20-30B. Considering AMD was immediately selling every chip they made.. *before* they started getting systems with their chips inside offered by the likes of Dell.. I find it hard to believe their bottom line would've been improved that much if that of which Intel is accused didn't occur.

Actually, it was hugely significant...
If you go back and look at AMD's GM prior to the rebate program and after the rebate program, you will see that while they were selling a large unit marketshare, their dollar marketshare dropped massively...even though their performance was actually increasing.
If AMD had maintained their GM during that period, they could easily have made an extra $10 Billion or more.

Since by law all anti-trust damages are tripled, that would be $30 billion+...

In fact, AMD commisioned a study on it early last year, and it showed (IIRC) ~$60 Billion in damages to the public (though it should be kept in mind that this was an AMD commisioned study).

 

[VirtualLarry]

Originally posted by: Yoxxy
EU hates monopolies.

Monopolies can be good for society.

Tell that to the economy and society of the former USSR. 'Nuf said.

 

[The-Noid]

Government controlled monopolies i.e communism is a completely different concept than monopolies in a single industry.

 

[Idontcare]

Originally posted by: CTho9305

Why didn't AMD have Charter make chips befor C2D was released.

That agreement was made back in 2004. Last time I checked, 2004 was before 2006, which would mean it wasn't motivated by Core 2. Somebody might want to check my math though .

I believe your math is correct but more importantly the technology node Chartered had at the time was not generating top-performing xtors and was never intended to.

Chartered was to take up the low-power mobile processor and low bin-speed sempron volume, a perfect fit for the low-power low-performance xtors comprising Chartered's technology.

 

[zsdersw]

Originally posted by: Viditor
Actually, it was hugely significant...
If you go back and look at AMD's GM prior to the rebate program and after the rebate program, you will see that while they were selling a large unit marketshare, their dollar marketshare dropped massively...even though their performance was actually increasing.
If AMD had maintained their GM during that period, they could easily have made an extra $10 Billion or more.

So *all* of that change in GM is Intel's fault? I doubt it. What about the broader economy and its influence?

In fact, AMD commisioned a study on it early last year, and it showed (IIRC) ~$60 Billion in damages to the public (though it should be kept in mind that this was an AMD commisioned study).

Damages to the public? Are you kidding? The public couldn't care less how much AMD makes on each chip.. a dip in GM means nothing to the consumer.

 

[Idontcare]

Originally posted by: zsdersw

In fact, AMD commisioned a study on it early last year, and it showed (IIRC) ~$60 Billion in damages to the public (though it should be kept in mind that this was an AMD commisioned study).

Damages to the public? Are you kidding? The public couldn't care less how much AMD makes on each chip.. a dip in GM means nothing to the consumer.

The damages to the public are intended to reflect the lack of corresponing declines in Intel's GM, and as it is the consumer who pays for those GM's the interpretation by the third party audit group (albeit commissioned by AMD) is that the consumer was unwittingly fleeced by Intel in Intel's pursuit of maintaining their higher GM in the face of technologically competent products from AMD.

Not saying the attribution is correct, just giving you the background on what the spirit of the "public damage" message means.

But you do not want to be too rash to minimize the issue the DOJ will have with this information. They will confirm it of course with their own economists, but it is the kind of "proof of the pudding is in the eating" type evidence of the existance of a monopoly abusing their powers.

Sustaining irrationally high GM's in a market which is quite sensitive to product performance will raise eyebrows, doing it when you have 80+% of the market share will raise more than just a few eyebrows.

 

[Viditor]

Originally posted by: Idontcare

Originally posted by: zsdersw

In fact, AMD commisioned a study on it early last year, and it showed (IIRC) ~$60 Billion in damages to the public (though it should be kept in mind that this was an AMD commisioned study).

Damages to the public? Are you kidding? The public couldn't care less how much AMD makes on each chip.. a dip in GM means nothing to the consumer.

The damages to the public are intended to reflect the lack of corresponing declines in Intel's GM, and as it is the consumer who pays for those GM's the interpretation by the third party audit group (albeit commissioned by AMD) is that the consumer was unwittingly fleeced by Intel in Intel's pursuit of maintaining their higher GM in the face of technologically competent products from AMD.

Not saying the attribution is correct, just giving you the background on what the spirit of the "public damage" message means.

But you do not want to be too rash to minimize the issue the DOJ will have with this information. They will confirm it of course with their own economists, but it is the kind of "proof of the pudding is in the eating" type evidence of the existance of a monopoly abusing their powers.

Sustaining irrationally high GM's in a market which is quite sensitive to product performance will raise eyebrows, doing it when you have 80+% of the market share will raise more than just a few eyebrows.

That was one of the best summations I have read anywhere...nice job IDC!

 

[Acanthus]

I wouldnt place AMDs pitfalls solely on Intel being anti-competitive.

Phailnom barely beats its predecessor, the ATI aquisition was horrible timing, their marketing has been abysmal for years, they fail to outsource when they cant meet demands of OEMs... and the list goes on and on...

I will admit that it likely contributed to their current situation, but not $30b worth.

 

[Idontcare]

Originally posted by: Acanthus
I wouldnt place AMDs pitfalls solely on Intel being anti-competitive.

Phailnom barely beats its predecessor, the ATI aquisition was horrible timing, their marketing has been abysmal for years, they fail to outsource when they cant meet demands of OEMs... and the list goes on and on...

I will admit that it likely contributed to their current situation, but not $30b worth.

We should temper our expectations of what is reasonable product-line to product-line improvements based on the R&D resources available to make those improvemtns.

Consider, by whatever metric pleases you, the improvement Intel was able to generate in their Netburst -> Core transition. Normalize this IPC improvement in context of the financial investment Intel made to drive that improvement.

Now consider, by same metrics of course, the improvement AMD was able to generate in their K8 -> K10 transition. Normalize this IPC improvement in contect to the financial investment AMD made to drive that improvement.

Now compare the ROI (return on investment) each company derived from their investments. There are second-order effects (laws of diminishing returns, etc) which will prevent a linear ROI per unit investment scaling, of course, but to first-order this ought to be a linear scaling.

Phailnom would be a reasonable assertion if AMD generated Phenom type results with Intel type investments. But that was simply not the situation.

I find it hard to fault AMD for their product SKU's relative to Intel, this is pure numerology insomuch as it can ever be. Intel invested more and got more in return for that investment. QED.

Now why was Phenom not a more impressive product? Well AMD's contention could be that by having a legitimate business model be undermined by an abusive monopoly tactic they were starved of the GM's and thusly the R&D dollars needed 2-3 years ago to invest into a larger development team.

Ergo from there as logic dictates.

 

[Viditor]

Originally posted by: Acanthus
I will admit that it likely contributed to their current situation, but not $30b worth.

To be clear here...the $30 billion number is the award, not the amount of damages.
In an anti-trust lawsuit, awards are different than in other lawsuits. It actually states in the statute that awards are to be triple the amount of the damages (and I believe that this the only time that treble damages are prescribed by law). Of course it also doesn't allow awarding of expenses (legal, flights, data collection, etc...) as other lawsuits do...I guess they figure treble damages is enough.

 

[Acanthus]

Originally posted by: Idontcare

Originally posted by: Acanthus
I wouldnt place AMDs pitfalls solely on Intel being anti-competitive.

Phailnom barely beats its predecessor, the ATI aquisition was horrible timing, their marketing has been abysmal for years, they fail to outsource when they cant meet demands of OEMs... and the list goes on and on...

I will admit that it likely contributed to their current situation, but not $30b worth.

We should temper our expectations of what is reasonable product-line to product-line improvements based on the R&D resources available to make those improvemtns.

Consider, by whatever metric pleases you, the improvement Intel was able to generate in their Netburst -> Core transition. Normalize this IPC improvement in context of the financial investment Intel made to drive that improvement.

Now consider, by same metrics of course, the improvement AMD was able to generate in their K8 -> K10 transition. Normalize this IPC improvement in contect to the financial investment AMD made to drive that improvement.

Now compare the ROI (return on investment) each company derived from their investments. There are second-order effects (laws of diminishing returns, etc) which will prevent a linear ROI per unit investment scaling, of course, but to first-order this ought to be a linear scaling.

Phailnom would be a reasonable assertion if AMD generated Phenom type results with Intel type investments. But that was simply not the situation.

I find it hard to fault AMD for their product SKU's relative to Intel, this is pure numerology insomuch as it can ever be. Intel invested more and got more in return for that investment. QED.

Now why was Phenom not a more impressive product? Well AMD's contention could be that by having a legitimate business model be undermined by an abusive monopoly tactic they were starved of the GM's and thusly the R&D dollars needed 2-3 years ago to invest into a larger development team.

Ergo from there as logic dictates.

I would argue that unless AMDs research budget is so small that they simply didnt do any, there is no reason that they could produce a CPU that fails to significantly outperform its last generation (or in this case, 2 generations since K9 was dropped).

 

[Acanthus]

Originally posted by: Viditor

Originally posted by: Acanthus
I will admit that it likely contributed to their current situation, but not $30b worth.

To be clear here...the $30 billion number is the award, not the amount of damages.
In an anti-trust lawsuit, awards are different than in other lawsuits. It actually states in the statute that awards are to be triple the amount of the damages (and I believe that this the only time that treble damages are prescribed by law). Of course it also doesn't allow awarding of expenses (legal, flights, data collection, etc...) as other lawsuits do...I guess they figure treble damages is enough.

I didnt know that. Thanks

 

[foghorn67]

Originally posted by: Idontcare

Originally posted by: Acanthus
I wouldnt place AMDs pitfalls solely on Intel being anti-competitive.

Phailnom barely beats its predecessor, the ATI aquisition was horrible timing, their marketing has been abysmal for years, they fail to outsource when they cant meet demands of OEMs... and the list goes on and on...

I will admit that it likely contributed to their current situation, but not $30b worth.

We should temper our expectations of what is reasonable product-line to product-line improvements based on the R&D resources available to make those improvemtns.

Consider, by whatever metric pleases you, the improvement Intel was able to generate in their Netburst -> Core transition. Normalize this IPC improvement in context of the financial investment Intel made to drive that improvement.

Now consider, by same metrics of course, the improvement AMD was able to generate in their K8 -> K10 transition. Normalize this IPC improvement in contect to the financial investment AMD made to drive that improvement.

Now compare the ROI (return on investment) each company derived from their investments. There are second-order effects (laws of diminishing returns, etc) which will prevent a linear ROI per unit investment scaling, of course, but to first-order this ought to be a linear scaling.

Phailnom would be a reasonable assertion if AMD generated Phenom type results with Intel type investments. But that was simply not the situation.

I find it hard to fault AMD for their product SKU's relative to Intel, this is pure numerology insomuch as it can ever be. Intel invested more and got more in return for that investment. QED.

Now why was Phenom not a more impressive product? Well AMD's contention could be that by having a legitimate business model be undermined by an abusive monopoly tactic they were starved of the GM's and thusly the R&D dollars needed 2-3 years ago to invest into a larger development team.

Ergo from there as logic dictates.

AMD's press releases and marketing gimmicks make it easy to fault AMD for Phailnom.

 

[Viditor]

Originally posted by: Acanthus

I would argue that unless AMDs research budget is so small that they simply didnt do any, there is no reason that they could produce a CPU that fails to significantly outperform its last generation (or in this case, 2 generations since K9 was dropped).

This has happened to both Intel and AMD before though...
For example, the P4 performed significantly worse than the P3 for the beginning of it's life.
The K6-III wasn't as good as the K6-II.

The Phenom has only been out for a few months...we should wait at least a few more quarters before declaring the entire design a lost cause.

 

[Idontcare]

Originally posted by: Acanthus
I would argue that unless AMDs research budget is so small that they simply didnt do any, there is no reason that they could produce a CPU that fails to significantly outperform its last generation (or in this case, 2 generations since K9 was dropped).

You would make the argument that the first quanta of ROI to be garnered in transitioning from zero investment to something !=0 is the creation of a significantly outperforming CPU?

 

[zsdersw]

What is VIA's CPU design budget? What is AMD's?

How can VIA design a CPU that's pretty good for what it's aimed at?

 

[Nemesis 1]

Quote Modelworks

Not really.
Via has not been in a position to really be considered a competitor.
It would be like Toyota taking Porsche to court. They are both in the car market, but target different customers.

In your mind what would constitute a compeitive position? All companies have to start somewhere.

Via is aimed at the low end, average performance pc market. They aren't trying to be the cpu for gaming or number crunching. They remind me a lot of cyrix. Except cyrix wanted it all. Cyrix has ties to via and amd.

Do you work for Via? As you seem to know . What markets Via is targeting now and in the future. As I said all companies have to start somewhere. IF company A has hugh Market share and companyB sells its products at a price point that effectively keeps company C from entering the Market in other areas.

Quote CTho9305


At first glance, the deal doesn't look that evil - it looks a lot like like an ordinary volume discount. However, when you look again, you can see that it's set up in a way that makes it impossible for competitors to enter the market.

At AMDs curent prices doesn't this apply to Via also. Or is it just AMD we are concerned with.

At first glance, the deal doesn't look that evil

Evil! One thing I like about offtopic section and Politics is you get to know other posters.

What is your positition on gay rights and abortion? Would you consider these evil or a Choice thing?

IF before AMD64. AMD had zero server market . Than according to some here Intel couldn't do anything wrong. Because AMD wasn't in thart market . Thats what your saying about VIA. YOU can't have it both ways guys.

 

[steve1616]

Originally posted by: Yoxxy

Originally posted by: waffleironhead

Originally posted by: Yoxxy


Generally if you want to use the theory of oligopolies, Intel is already a monopoly by taking more than 80% market share in a duopoly. (YES VIA/IBM do take a small share).

There are also barriers to entries because of the X86 license.

You are confusing "Pure Monopoly" with "Monopolistic (Monopoly) Competition"

Monopoly does not mean only one player in a single market.

Im refering to monopoly as

Monopoly: Exclusive control by one group of the means of producing or selling a commodity or service

Repeat: Intel is not a Monopoly.

Repeat Intel is a Monopoly, why do you think Anti-Trust is coming from the State of New York and EC, just for fun?

Again read up on Pure Monopoly and the definition of Monopoly.

Intel is a Monopoly...

I have never said I disprove of Monopolies, as barring the salt monopolies of the 17th Century they are actually quite good for society. As they become lax in their production of new and innovative goods and services there is always another power that comes into market share.

<--- Graduated Summa in Econ.

Yoxxy, don't your statements disprove each other. You say that monopolies are quite good for society. Yet you also state that the monopoly has to become lax in their production of goods and services for another power to come into market share. Wouldn't it always be better to just have good competition where neither business can afford to become lax.
Government was formed for many reasons, but one of the main reasons was to prevent, or at least regulate monopolies. Sometimes monopolies can't be avoided, so the government is suppose to do their best to regulate them.
I can give you a ton of examples to where monopolies could destroy society. Yes, just monopolies. Also, please don't give me your pure monopoly speech, because I do know the difference between a pure monopoly and a monopoly.
Monopolies are always very bad. I am not saying that it is always a business fault if they happen to be a monopoly, but that it is never good for society. Cargil controls most of the grain in the world, and they are a monopoly. Not a pure monopoly, but a monopoly. Yet, if they so desired, they could make everybody in all countries starve. The government regulates them so this doesn't happen. Point being, I would like you to find me one example of a monopoly being good for a society.

 

[VirtualLarry]

Originally posted by: Yoxxy
Government controlled monopolies i.e communism is a completely different concept than monopolies in a single industry.

Lack of choice/competitions in both cases causes the market to stagnate. What about that concept don't you get?

 

[Nemesis 1]

Heres is a chart notice intels sharp market drop . notice who gained . It wasn't AMD because they couldn't produce anymore chips. It was others that gained . Than notice when AMD got Fab 36 was finely completed a year late. Its market share jumped . In the US courts Intels Laywers will eat AMD lawyers alive. What happens in the EVIL EU who knows or cares. Post 36.


http://www.xtremesystems.org/f...ad.php?t=176878&page=2

 

[Acanthus]

Originally posted by: Viditor

Originally posted by: Acanthus

I would argue that unless AMDs research budget is so small that they simply didnt do any, there is no reason that they could produce a CPU that fails to significantly outperform its last generation (or in this case, 2 generations since K9 was dropped).

This has happened to both Intel and AMD before though...
For example, the P4 performed significantly worse than the P3 for the beginning of it's life.
The K6-III wasn't as good as the K6-II.

The Phenom has only been out for a few months...we should wait at least a few more quarters before declaring the entire design a lost cause.

The P4 was a radically different design from the P3, and was designed specifically to scale to very high clockspeeds with a very long pipeline.

While i agree to come extent that we cant call the entire architecture a failure, Phenom is very heavily based on Athlon-64 technology and doesnt have the same "radical changes" that the P4 did.

AMD hasnt been hinting at anything like the CPU scaling well, and there doesnt seem to be lots of headroom for clocking the chip up. Heres hoping i am wrong and B3 or C0 will be a radical change in that respect.