Intel's agreement with the FTC

[Yousuf Khan]

http://download.intel.com/pressroom/legal/ftc/FTC_Final_Executed_Agreement.pdf

***
So far, I've read that Intel-FTC agreement requires:

(1) Intel will allow competitors to use any foundry they like.
(2) Intel will not sue a competitor for a year, if that competitor gets
sold to a third party, until a new cross license can be negotiated.
(3) VIA gets an extension to its Intel license.
(4) Intel cannot price a chipset/cpu combo below the the price of the
CPU alone.
(5) Intel cannot offer discounts based on market share levels.
(6) Intel can only offer volume discounts.
(7) Intel cannot punish a customer for having a relationship with a
competitor.
(8) Intel can enter into an exclusivity agreement with a customer, only
so long as is required to recoup capital investment for that customer (I
think this may refer to Apple). It cannot enter into more than 10 such
agreements at a time.
(9) Intel cannot give extraordinary lump sum assistance to customers (Dell).
(10) Intel cannot degrade the performance of its products in the
presence of a competitor's product (GPUs). Unless it's a bug.
(11) Intel has to provide an interface roadmap to its competitors.
(12) Intel must clearly state that its compilers will not optimize for
its competitors' products. It will not be allowed to misrepresent this
anymore.
(13) It has to reimburse its compiler customers who thought this
compiler would work with compatible x86 processors, based on Intel's
misrepresentations.
(14) Intel has to reveal that the SYSmark and MobileMark benchmarks are
optimized for Intel processors only.

 

[Robert Myers]

http://download.intel.com/pressroom/legal/ftc/FTC_Final_Executed_Agre...

***
So far, I've read that Intel-FTC agreement requires:

(1) Intel will allow competitors to use any foundry they like.
(2) Intel will not sue a competitor for a year, if that competitor gets
sold to a third party, until a new cross license can be negotiated.
(3) VIA gets an extension to its Intel license.
(4) Intel cannot price a chipset/cpu combo below the the price of the
CPU alone.
(5) Intel cannot offer discounts based on market share levels.
(6) Intel can only offer volume discounts.
(7) Intel cannot punish a customer for having a relationship with a
competitor.
(8) Intel can enter into an exclusivity agreement with a customer, only
so long as is required to recoup capital investment for that customer (I
think this may refer to Apple). It cannot enter into more than 10 such
agreements at a time.
(9) Intel cannot give extraordinary lump sum assistance to customers (Dell).
(10) Intel cannot degrade the performance of its products in the
presence of a competitor's product (GPUs). Unless it's a bug.
(11) Intel has to provide an interface roadmap to its competitors.
(12) Intel must clearly state that its compilers will not optimize for
its competitors' products. It will not be allowed to misrepresent this
anymore.
(13) It has to reimburse its compiler customers who thought this
compiler would work with compatible x86 processors, based on Intel's
misrepresentations.
(14) Intel has to reveal that the SYSmark and MobileMark benchmarks are
optimized for Intel processors only.

You left out the part where the federal government is dictating
details of technology (PCI-X for six years).

Good news for fanboys. For the industry? For real consumers? What a
joke.

Robert.

 

[Robert Myers]

You left out the part where the federal government is dictating
details of technology (PCI-X for six years).

Good news for fanboys.  For the industry?  For real consumers?  What a
joke.

http://news.cnet.com/8301-13924_3-20012636-64.html

"We cannot simply assume that the settlement equates to a victory for
consumers," Joshua Wright, an assistant professor of law at George
Mason University Law School in Arlington, Va., wrote in a blog post
after Wednesday's decision. U.S. government agency settlements are not
always that "meaningful, from a consumer welfare perspective,"
according to Wright, who also served as a scholar in residence at the
FTC.

Robert.

 

[Intel Guy]

Robert Myers full-quoted:

http://news.cnet.com/8301-13924_3-20012636-64.html

"We cannot simply assume that the settlement equates to a victory
for consumers,"

Would doing nothing and allowing Intel to continue to strong-arm
customers have been better for consumers?

 

[Intel Guy]

You left out the part where the federal government is dictating
details of technology (PCI-X for six years).

Good news for fanboys. For the industry? For real consumers?
What a joke.

The joke is that PCIe was foisted on consumers as a replacement for AGP
primarily to drive redundant video card and motherboard sales when the
reality was that there was a negligible real-world performance increase
with the new bus.

Do you really want to keep seeing needless forced-obsolescence for your
investment in computing hardware?

Intel should be forced to support PCIe for 12 years - not 6.

 

[Sebastian Kaliszewski]

You left out the part where the federal government is dictating
details of technology (PCI-X for six years).

So what?

Good news for fanboys. For the industry? For real consumers? What a
joke.

Yes of course, monopolistic practices are good for customes -- according
to Robert Myers

rgds
\SK

 

[Robert Myers]

Robert Myers full-quoted:



Would doing nothing and allowing Intel to continue to strong-arm
customers have been better for consumers?

I should have quoted the rest:

<quote>

Wright sees other potential problems. "This settlement has the FTC
getting itself involved in Intel's business arrangements, competitive
strategy, and even product design at a remarkably deep level," he
said, expressing concern about government micromanagement of Intel
business practices.

</quote>

It's a very bad precedent and, yes, it would have been better to have
done nothing, if the something included having bureaucrats and lawyers
micromanage a business, especially a high-technology business.

Robert.

 

[Robert Myers]

So what?


Yes of course, monopolistic practices are good for customes -- according
to Robert Myers

And you don't think the FTC and the European Commission are
monopolies? Really, Sebastian, what's written in your history books
about the miracles of government-planned businesses and all the
munificent benefits they bestow through the intrusive exercise of
power?

Robert.

 

[Rick Jones]

The joke is that PCIe was foisted on consumers as a replacement for
AGP primarily to drive redundant video card and motherboard sales
when the reality was that there was a negligible real-world
performance increase with the new bus.

And how about today with PCIe Gen2? I'm not a HW guy, but how much
headroom remained in AGP, or PCI-X for that matter?

Do you really want to keep seeing needless forced-obsolescence for
your investment in computing hardware?

Intel should be forced to support PCIe for 12 years - not 6.

Would you like ISA bus support with that?

Given that PCIe Gen1 seems to date to 2004 and it is now 2010, if
Intel supports PCIe for another 6 years, it will indeed have supported
it for 12 years.

IIRC Infiniband is already at 40 Gbit/s, which probably makes for
"fun" with a dual-port HCA even in a x8 PCIe Gen2 slot. Dual-port 40
Gbit Ethernet will be equally fun, and haven't the IEEE done the spec
for 100Gbit now (or am I getting ahead of them?)

rick jones

* not clear if first product shipped then or if that was just the spec
- I'm going from a Wikipedia article - btw, wikipedia asserted that
AGP was from 1997

 

[Robert Myers]

As I said, those were my initial quick reads. But the PCIe issue is
something they did for Nvidia, to give it plenty of time to adjust. Also
it falls into the general idea of Intel now having to provide product
roadmaps to its competitors. This is generally good because Intel has
very recently been playing games with the USB 3.0 specs, first delaying
giving them the specs, and then when it finally gave them the specs, it
decided that it might want to replace it entirely with some kind of
optical link instead. Does nothing but provide instability in the PC
standards.


Probably really good news for everybody listed above, including you Robert.

I have yet to read that any of Intel's actions caused harm to
consumers (from anyone credible).

As to Intel playing interconnect games, so what else is new?

One of the ironies here is that if Intel *did* keep prices
"artificially high," it would have benefited AMD, who has a hard time
selling chips at a profit.

As to good news for me, I don't see any. A regulatory tax on Intel's
business. More obstacles to innovation. Holding on to PCI-X is *not*
good news.

Fortunately, because of its ruthless business tactics, Intel can just
throw money at things, which means we will soon see photons as a
bigger part of the mix, and not a moment too soon, in spite of the
government (and AMD) interference.

Robert.

 

[Yousuf Khan]

I have yet to read that any of Intel's actions caused harm to
consumers (from anyone credible).

5 anti-trust lawsuits with $2.7 billion dollars in fines and
compensation later, serves adequately as a credible witness.

As to Intel playing interconnect games, so what else is new?

Yes, so what else is new? Well namely it's now going to stop, that's
what's new.

One of the ironies here is that if Intel *did* keep prices
"artificially high," it would have benefited AMD, who has a hard time
selling chips at a profit.

The low prices you've seen up until now are nothing compared to what
will happen from now on.

With more marketshare AMD will now be able to afford the lower prices
without going into a loss. That's what will make it different.

As to good news for me, I don't see any. A regulatory tax on Intel's
business. More obstacles to innovation. Holding on to PCI-X is *not*
good news.

Oh boo-hoo Robert, why do you care what regulations are put on Intel?
And not everything is a tax, get that straight Robert. There are taxes,
and then there are regulations, two totally different functions of
government. No taxes here. Enough of the civics lessons.

And why do you care if PCI-e stays around another few years or not? It
was bound to stay anyways.

Fortunately, because of its ruthless business tactics, Intel can just
throw money at things, which means we will soon see photons as a
bigger part of the mix, and not a moment too soon, in spite of the
government (and AMD) interference.

Robert.

If you think Intel was going to be ready with a photonic interface as
anything other than a red-herring to throw the PC market place into
chaos, then you're ready to buy the Brooklyn bridge. Intel has played
these standards games before and it'll play them again if it wasn't put
on a leash.

Yousuf Khan

 

[Yousuf Khan]

The joke is that PCIe was foisted on consumers as a replacement for AGP
primarily to drive redundant video card and motherboard sales when the
reality was that there was a negligible real-world performance increase
with the new bus.

Actually, as I remember it, PCI-e was foisted on the consumers to avoid
them adopting AMD's Hypertransport as a standard. When AMD developed HT,
Intel had no answer to it for nearly 8 years. So it threw the
red-herring of a next generation, serial PCI in as the answer. AMD
didn't object, as it wasn't really a competitor to HT, and AMD itself
could use it. Video cards that could connect directly through HT
would've actually been much faster than PCI-e or AGP, since there would
a much smaller overhead, but it would've been proprietary to only AMD
systems as Intel would've never adopted it, even if it was free.

Do you really want to keep seeing needless forced-obsolescence for your
investment in computing hardware?

Intel should be forced to support PCIe for 12 years - not 6.

At least the version of PCIe available by then.

Yousuf Khan

 

[Robert Myers]

5 anti-trust lawsuits with $2.7 billion dollars in fines and
compensation later, serves adequately as a credible witness.

Who asked Intel's opinion? Everyone in the business press says that
consumers have not been harmed. If the FTC could have made such a
charge stick, it would have done so. What do I care what happens if
consumers haven't been harmed?

Yes, so what else is new? Well namely it's now going to stop, that's
what's new.

They won't stop playing interconnect games.

The low prices you've seen up until now are nothing compared to what
will happen from now on.

With more marketshare AMD will now be able to afford the lower prices
without going into a loss. That's what will make it different.

Wish I could lay a bet with you about market share. How much do you
think AMD's market share will grow? It will double? Then it will
start making money?

Oh boo-hoo Robert, why do you care what regulations are put on Intel?
And not everything is a tax, get that straight Robert. There are taxes,
and then there are regulations, two totally different functions of
government. No taxes here. Enough of the civics lessons.

Everything a business does is paid by the end consumer. You can say
it's not a tax, if you like, but the costs come out of your pocket.
Well, mine, since you don't buy Intel products.

And why do you care if PCI-e stays around another few years or not? It
was bound to stay anyways.


If you think Intel was going to be ready with a photonic interface as
anything other than a red-herring to throw the PC market place into
chaos, then you're ready to buy the Brooklyn bridge. Intel has played
these standards games before and it'll play them again if it wasn't put
on a leash.

I don't know enough details to make a prediction, and I don't think
you do, either. What's been coming out of Intel looks pretty real to
me.

Robert.

 

[Yousuf Khan]

Who asked Intel's opinion? Everyone in the business press says that
consumers have not been harmed. If the FTC could have made such a
charge stick, it would have done so. What do I care what happens if
consumers haven't been harmed?

Who asked Intel's opinion? Probably all of the judges and lawyers in
those various court cases it lost, I would guess. Some of those lawyers
belonged to Intel, so at the very least they should have.

Everyone in the business press says that consumers have not been harmed.

Ah, if only the business press were the ones judging Intel, then you
would've been completely right. But since they weren't, so you're
completely wrong.

I mean who could be more impartial than writers whose salaries depend
entirely on advertising revenue from big advertisers, like Intel?

If the FTC could have made such a charge stick, it would have done so.

Have you actually read the PDF? That's pretty much all that the FTC
keeps saying, it's within pretty much every section start. Pretty much
every section has something about "Benefit to Customer or End User",
which means it's talking about "harm to consumers".

Put down your pink sunglasses, the company you love so much has been
found guilty of everything that it has been accused of for so many years.

They won't stop playing interconnect games.

There's a piece of paper with an agreement with the FTC that says they
will. Well, actually there's several pieces of paper, they also had an
agreement with AMD a few months back on which this FTC agreement is
modelled on which also says similar things.

Wish I could lay a bet with you about market share. How much do you
think AMD's market share will grow? It will double? Then it will
start making money?

According to Intel's previous agreement with AMD, Intel will allow AMD
upto 35% before it starts the games again, er, I mean before Intel deems
its agreement with AMD to have been fulfilled.

Everything a business does is paid by the end consumer. You can say
it's not a tax, if you like, but the costs come out of your pocket.
Well, mine, since you don't buy Intel products.

If that's your screwed up definition of a tax, then we've just been
given a tax break.

I don't know enough details to make a prediction, and I don't think
you do, either. What's been coming out of Intel looks pretty real to
me.

They'll be real alright, but probably in another decade or so. Intel is
showing off science projects hoping people will wait for them to become
real products and ignore actual current solutions. Remember the wireless
laptop recharging system that Intel showed off in an IDF a couple of
years back? That's otherwise known as a Tesla coil, and looks like it's
something that came out of a steampunk design from the 19th century.
Intel is just entertaining people, not really providing real solutions
to anything.

Yousuf Khan

 

[Sebastian Kaliszewski]

And you don't think the FTC and the European Commission are
monopolies?

As any other government agencies. Government has monopoly by its very
nature.

Really, Sebastian, what's written in your history books
about the miracles of government-planned businesses and all the
munificent benefits they bestow through the intrusive exercise of
power?

And what's written in qyour history books about miracles of many
(competing) goverments over the same territory?

Or if you didn't notice neither EC nor FTC run business.

\SK

 

[Sebastian Kaliszewski]

I should have quoted the rest:

<quote>

Wright sees other potential problems. "This settlement has the FTC
getting itself involved in Intel's business arrangements, competitive
strategy, and even product design at a remarkably deep level," he
said, expressing concern about government micromanagement of Intel
business practices.

</quote>

It's a very bad precedent and, yes, it would have been better to have
done nothing, if the something included having bureaucrats and lawyers
micromanage a business, especially a high-technology business.

Oh. If you didn't notice it's a settelemnt. IOW Intel agreed to that
conditions. Intel still had a choice to go to court -- if everything was
all right as you portray they should fear not. Or, maybe the chance of
unfavourable judgement was real, as not everything was all right?

\SK

 

[Robert Myers]

They'll be real alright, but probably in another decade or so. Intel is
showing off science projects hoping people will wait for them to become
real products and ignore actual current solutions. Remember the wireless
laptop recharging system that Intel showed off in an IDF a couple of
years back? That's otherwise known as a Tesla coil, and looks like it's
something that came out of a steampunk design from the 19th century.
Intel is just entertaining people, not really providing real solutions
to anything.

If there are actual "solutions," they are of limited usefulness. The
ability of processors to crunch data has begun to stress the ability
of wires even theoretically to deliver data to them fast enough.
GPU's used for computation will make the situation even worse, as will
the planned addition of vector processors to x86.

The world you imagine for computers is a disappearing mirage. The
concentration of capital required to advance the technology means that
there will be fewer players, and, aside from token competition,
essentially only one player, at least for now.

From my perspective, there is really only one problem left: how to get
the data there fast enough. What's left of processor design is how to
find ways to tolerate (hide) even more latency than processors now
can. The rest of the problem is all about interconnect, and the
solution isn't going to come in the form of electrons.

If AMD (or IBM/AMD) has the resources to move us forward on those
issues, then I'm interested. If not, then, as far as I'm concerned,
it's only an internecine struggle that wastes resources to no one's
benefit.

I really couldn't care less about how cheaply you can buy processors
for your next home-brew experiment, and I don't think the government
should, either.

At the somewhat less-than-cutting-edge, the story is probably quite
different. Most people can't use the muscle that is available at
affordable prices now, and there will be lots of players who can
produce "good enough" processors that don't rely on licenses from
Intel.

Were I Intel, I'd be *much* more worried about ARM than about AMD.
That doesn't even account for the fact that China has both the capital
and the engineering expertise to do more or less whatever it wants,
independent of both Intel *and* ARM.

Aside from optical interconnects, this entire thread has been about
yesterday's news.


Robert.

 

[Robert Myers]

Windows requires a PCI bus so that alone will keep from going anywhere.

But not as a x16 slot that nVidia can plug its cards into.

Otherwise, why am I paying government lawyers to reserve space on
every motherboard I buy for at least six years?

Robert.

 

[Yousuf Khan]

But not as a x16 slot that nVidia can plug its cards into.

Otherwise, why am I paying government lawyers to reserve space on
every motherboard I buy for at least six years?

Robert.

It hasn't said that you need to keep the slots around, just the bus.
That means GPUs can be soldiered onto motherboards using PCIe lines
directly.

Yousuf Khan

 

[Yousuf Khan]

On 8/6/2010 2:21 PM, Robert Myers wrote:

Were I Intel, I'd be *much* more worried about ARM than about AMD.
That doesn't even account for the fact that China has both the capital
and the engineering expertise to do more or less whatever it wants,
independent of both Intel *and* ARM.

Well, of course, that's because the ARM market is the only market that
Intel hasn't been able to bring to brink of near destruction with its
anti-competitive actions. Remember it wasn't so long ago that Intel
itself was selling ARM processors, in the form of Xscale. Nobody wanted
to buy them from Intel, because everybody was afraid of dealing with
Intel. Instead they bought their ARM's from anybody else but Intel. It's
not that there was anything wrong with the Xscales, in fact they started
selling like hotcakes -- once they were produced by Marvell instead.
Nobody wants to deal with the mafia if they have a choice.

Aside from optical interconnects, this entire thread has been about
yesterday's news.

What this is supposed to be a form of derision from you? News is always
about yesterday's news.

Even optical interconnects are yesterday's news. Why not just wait for
quantum interconnects?

Yousuf Khan