CCI completes probe into Intel malpractice [The Times of India]

[galego]

Intel has been under investigation during last 18 months for bad practices.

Submission to the DG is made in the cases "where the Commission has found prima facie evidence of violating competition norms".

http://articles.timesofindia.indiat...tition-commission-trade-regulator-competition

:whiste:

Because SEVERAL of you have reported this thread now: this is a legitimate and factual news article, properly quoted and attributed. You may not like the poster nor the contents, but per our rules this is a legitimate thread and is relevant to this forum.
-ViRGE

 

[videogames101]

this guy is getting ridiculous

 

[SOFTengCOMPelec]

Seemingly, almost everyday, google is hitting the headlines because they broke some privacy law somewhere, or google glass is annoying people, or they were avoiding taxes in that country, or they have not complied with an EU court order to delete some data within 35 days, etc etc etc etc etc.

It's a fact of life, massively big computer corporations (Intel, Google, Apple etc) are often hitting the headlines.
Is this really news, worth posting here ?

 

[blackened23]

IS this a troll thread?

 

[galego]

Seemingly, almost everyday, google is hitting the headlines because they broke some privacy law somewhere, or google glass is annoying people, or they were avoiding taxes in that country, or they have not complied with an EU court order to delete some data within 35 days, etc etc etc etc etc.

It's a fact of life, massively big computer corporations (Intel, Google, Apple etc) are often hitting the headlines.
Is this really news, worth posting here ?

I think that this kind of material is important specially if accusations are verified. In my opinion it seems to be more relevant than other threads here. If other people disagrees they can ignore this thread just as I ignore others.

 

[SOFTengCOMPelec]

I think that this kind of material is important specially if accusations are verified. In my opinion it seems to be more relevant than other threads here. If other people disagrees they can ignore this thread just as I ignore others.

Intel (I think) have got into trouble for similar things before, in other countries. I was annoyed, when it transpired that Intel were fined for programming their special Intel compiler, to worsen (effectively slow down, or not optimally speed up) compiled code, if it detected a non-Intel processor (i.e. AMD).

This I consider particularly bad of Intel, but in all fairness, I would prefer to hear Intel's defence of the situation, before casting stones.

I have heard stories of how Intel has done bad things (competition wise), but it is never 100% apparent, what stories are true, which are exaggerated, and which are false.

 

[galego]

Intel (I think) have got into trouble for similar things before, in other countries. I was annoyed, when it transpired that Intel were fined for programming their special Intel compiler, to worsen (effectively slow down, or not optimally speed up) compiled code, if it detected a non-Intel processor (i.e. AMD).

This I consider particularly bad of Intel, but in all fairness, I would prefer to hear Intel's defence of the situation, before casting stones.

I have heard stories of how Intel has done bad things (competition wise), but it is never 100% apparent, what stories are true, which are exaggerated, and which are false.

When I revealed here in the forums that Intel cheated its compiler, at first a bunch of people started attacking me, then some other checked the claims, verified that the claims were true, and finally shared their frustration because nobody informed them of such practices, before I did.

I am merely reporting a news about Intel being investigated once again for bad practices. Each one can take their own conclusion. This news can be unimportant for some people but important for others.

 

[Phynaz]

Somebody has too much time on their hands.

 

[SOFTengCOMPelec]

When I revealed here in the forums that Intel cheated its compiler, at first a bunch of people started attacking me, then some other checked the claims, verified that the claims were true, and finally shared their frustration because nobody informed them of such practices, before I did.

I've been amazed at the apparent silence on the Intel Compiler issue. It's especially surprising, given that I think they ended up being fined a huge amount of money, and also surprising that AMD did not widely publicize what happened.

Poorly educated wild guess at why AMD did not take it upon themselves to publicise what Intel did with the compiler:

Sometimes there is too much danger of tit-for-tat reprisals, so if AMD had "thrown the dirt" about what Intel had done, maybe Intel could have started throwing stuff (perhaps which we do not know about) about AMD, or done other things which AMD would have strongly disliked, or something.

Maybe the court(s) that were involved placed conditions on the fine (which I think was paid in compensation to AMD), restricting AMD's rights to publicize what happened.

The thing is, companies like AMD know that messing about too much with things like that, can cause critical meltdown situations, such as the relatively recent Samsung vs Apple "wars", and most people probably know how that has turned out.
i.e. Intel, as AMD's major, no1 competitor, could be especially nasty in the market place vs AMD, if Intel were wound up enough (N.B. my own take/guess/opinion) on the situation, probably best taken with a pinch of salt, while ignoring my post, 201%.

 

[A5]

The compiler stuff is a non-story that's been going on for at least as long as I've been around these forums. It isn't Intel's job to make their compiler work well on AMD chips.

Intel has repeatedly been hit for anti-competitive practices all around the world for just as long. It hasn't made any difference (except AMD got a $1B settlement in the last round in the US), and it isn't going to stop.

 

[Mallibu]

The whole world is conspiring against your favorite company. ^_^

 

[JQuilty]

It isn't Intel's job to make their compiler work well on AMD chips.

No, but it is certainly an abuse of market leader position to have their compiler intentionally cripple or take non-optimal paths on competitor's chips and then advertise benchmarks compiled with said compiler.

 

[SOFTengCOMPelec]

The compiler stuff is a non-story that's been going on for at least as long as I've been around these forums. It isn't Intel's job to make their compiler work well on AMD chips.

Intel has repeatedly been hit for anti-competitive practices all around the world for just as long. It hasn't made any difference (except AMD got a $1B settlement in the last round in the US), and it isn't going to stop.

N.B. I am NOT disagreeing with you here, it's just that I would prefer to know more about what really happened, before making my mind up.

I've wondered about exactly the same thing.
Is it that the compiler, simply is designed to do a good job for Intel cpu's, and will do what it can for other cpu's, such as AMD. In which case, that sounds quite reasonable (as you mentioned, in your post).

Or, does the Intel compiler actually "on purpose", worsen the code optimization because it is a non-Intel cpu.
I don't know enough about the issue to know the answer myself, without being able to examine the Intel compiler source code, and make my own mind up, or other sources of information, of 100% known reliability.

Also, you could simply say that it was up to AMD to produce their own optimized for AMD (and others, if necessary) compiler, which they did do, eventually, I believe.

 

[Mallibu]

"Winners make adjustments. Losers make excuses"
The compiler issue is another excuse for AMD shortcomings. What does prevent AMD from releasing their own compiler?
Why do developers use the Intel compiler (when they do), if it doesn't provide optimal performance for 1/4 of their user base?
And lastly, please provide a list of modern & popular software using the Intel compiler.

 

[galego]

The compiler stuff is a non-story that's been going on for at least as long as I've been around these forums. It isn't Intel's job to make their compiler work well on AMD chips.

That was never the issue, but I am not going into details in this thread. The real compiler issue was debated enough in another thread, with lots of useful information given.

Intel has repeatedly been hit for anti-competitive practices all around the world for just as long. It hasn't made any difference (except AMD got a $1B settlement in the last round in the US), and it isn't going to stop.

I don't know how to interpret this, really. It seems like a kind of justification given to Intel to do what they want. I hope this is an incorrect interpretation.

 

[Mallibu]

Taken from stackoverflow

Something about performance on AMD machines: It's not as good as Intel CPUs, but it's still way better than the MS C++ compiler (again, from my experience). The reason is that you can also target a generic CPU with SSE2 support (for example). Then AMD CPUs with SSE2 will not be discriminated much. Intel compiler on Intel CPU really steals the show, though. It's not all double rainbows and shiny unicorns, however. There have been some heavy accusations about binaries not-running at all on non-GenuineIntel CPUs and (this one is admitted) artificially induced inferior performance on CPUs by other vendors. Also note this is information from at least 3 years ago and it's validity as of now is unknown, BUT the new product descriptions gives binaries a carte blanche to run as slow as Intel sees fit on non-Intel CPUs.

I don't know what it is about Intel and why they make so good numeric computation tools, but have a look at this, too: http://julialang.org/. There is a comparison and if you look at the last row, MATLAB shines by defeating both C code and Julia, what strikes me is that the authors think the reason is Intel's Math Kernel Library.

I realize this sounds a lot like an advertisement for the Intel Compiler toolkit, but in my experience it really did the job well, and even simple logic dictates that the guys who make CPUs should know best how to program for them. IMO, the Intel C++ compiler squeezes every last bit of performance gain possible.

The news article was not about Intel's compilers. So neither will this thread. Please take that to another thread if you wish to continue
-ViRGE

 

[SOFTengCOMPelec]

Back on topic. If things like the initial quote at the top of this thread, is true, it could mean that, indirectly, Intel are conning computer users around the world into paying more money, than they should/need to for their computers, because they have hindered the competition, such as AMD.

Obviously, this is because cpu prices, may have ended up being higher than they need to be.

 

[blackened23]

No, but it is certainly an abuse of market leader position to have their compiler intentionally cripple or take non-optimal paths on competitor's chips and then advertise benchmarks compiled with said compiler.

Curious, I was under the impression that intel created x86 while AMD did business early on making exact clones of intel chips by reverse engineering them. :whiste:

So with that being the case, is it intel's duty to go through workarounds to ensure their software works on a competing product? x86 is their baby - intel has no such obligation to do anything of the sort. AMD just needs to step up as they did with their Athlon many years ago and stop making excuses non-stop. Less excuses, and better CPUs, end of story.

edit: this is off-topic, let's just please cease discussion here. (posted this before reading the entirety of thread)

 

[galego]

Back on topic. If things like the initial quote at the top of this thread, is true, it could mean that, indirectly, Intel are conning computer users around the world into paying more money, than they should/need to for their computers, because they have hindered the competition, such as AMD.

Obviously, this is because cpu prices, may have ended up being higher than they need to be.

I always believed that Intel chips are overpriced. Regarding the news the complaint was filled by a distributor and the allegations are that Intel abused its dominant market position and indulged in "restrictive trade practices".

 

[SOFTengCOMPelec]

I always believed that Intel chips are overpriced. Regarding the news the complaint was filled by a distributor and the allegations are that Intel abused its dominant market position and indulged in "restrictive trade practices".

I wish the courts, would let ME choose the punishment, for Intel, for allegedly doing this.

I would NOT fine them a single cent/penny, or restrict their activities in any way. They would NOT have to apologize to anyone, or change their marketing policies.
Just make a new law (for Intel), worldwide, which states that all future Intel cpu's, MUST be fully unlocked, including full access to all DISABLED cores (Socket 2011, etc) and ALL available on-chip features. Even on their cheapest cpu parts.

Intel are potentially getting too close to becoming a giant Monopoly, with AMD not doing so well, and Arm being some way behind (at least on desktop parts at the moment).
If that means that they charge SKY HIGH prices for their cpu's, it will be a sad day indeed, for many readers of these forums, and the rest of the world.

If these allegation(s) are true, I wish Intel would be a much more honourable company, instead of using "dirty tricks/tactics".

 

[JQuilty]

Curious, I was under the impression that intel created x86 while AMD did business early on making exact clones of intel chips by reverse engineering them. :whiste:

Half-correct. They did reverse-engineer the 8080. They got a license for x86 because Intel needed to have more than one supplier if they wanted to sell to IBM.

So with that being the case, is it intel's duty to go through workarounds to ensure their software works on a competing product?

It's not a workaround. There was a check to deliberately take a non-optimal code path if an Intel ID was not detected. There's no reason to do this other than to cripple competition.

x86 is their baby - intel has no such obligation to do anything of the sort.

And AMD created x86_64, which we use now. Relevance? None.

 

[beginner99]

If other people disagrees they can ignore this thread just as I ignore others.

If I had to ignore all threads you are trolling in I would have to leave anandtech...

 

[frozentundra123456]

I wish the courts, would let ME choose the punishment, for Intel, for allegedly doing this.

I would NOT fine them a single cent/penny, or restrict their activities in any way. They would NOT have to apologize to anyone, or change their marketing policies.
Just make a new law (for Intel), worldwide, which states that all future Intel cpu's, MUST be fully unlocked, including full access to all DISABLED cores (Socket 2011, etc) and ALL available on-chip features. Even on their cheapest cpu parts.

Intel are potentially getting too close to becoming a giant Monopoly, with AMD not doing so well, and Arm being some way behind (at least on desktop parts at the moment).
If that means that they charge SKY HIGH prices for their cpu's, it will be a sad day indeed, for many readers of these forums, and the rest of the world.

If these allegation(s) are true, I wish Intel would be a much more honourable company, instead of using "dirty tricks/tactics".

I "wish" I would win the lottery too. But in answer to your contentions, it is really breaking my heart to be able to pay less than 200 dollars for a quad core cpu that is infinitely faster than chips that used to cost 2 to 5 times as much. CPUs and computers in general are a terrific buy at this point in time. I dont see how anyone can consider prices "SKY HIGH" when you can get a cpu for 80.00 that is more than adequate for 90% of the users, and one for less than 200.00 that is adequate for all but the most high end uses.

 

[SOFTengCOMPelec]

I "wish" I would win the lottery too. But in answer to your contentions, it is really breaking my heart to be able to pay less than 200 dollars for a quad core cpu that is infinitely faster than chips that used to cost 2 to 5 times as much. CPUs and computers in general are a terrific buy at this point in time. I dont see how anyone can consider prices "SKY HIGH" when you can get a cpu for 80.00 that is more than adequate for 90% of the users, and one for less than 200.00 that is adequate for all but the most high end uses.

What I meant was, IF Intel continue down the road of becoming a MONOPOLY.

THEN
It may well cause Intel to charge "SKY HIGH" (if INTEL=MONOPOLY) prices for some/all of their cpu's.

Also, computers have been around for a long time, so even the $80 or $200 price points you mentioned, may be overly expensive (but I perfectly accept that, that is a much more contentious proposition).

EDIT: Arguably, if Intel were currently much less of a monopoly, e.g. If Steamroller (i.e. better competition) was with us now, and it was significantly better than the existing piledriver FX's, then maybe Haswell would be a lot less of a disappointment, for some people, in the enthusiasts market.
E.g. Solder, rather than TIM.

 

[frozentundra123456]

What I meant was, IF Intel continue down the road of becoming a MONOPOLY.

THEN
It may well cause Intel to charge "SKY HIGH" (if INTEL=MONOPOLY) prices for some/all of their cpu's.

Also, computers have been around for a long time, so even the $80 or $200 price points you mentioned, may be overly expensive (but I perfectly accept that, that is a much more contentious proposition).

EDIT: Arguably, if Intel were currently much less of a monopoly, e.g. If Steamroller (i.e. better competition) was with us now, and it was significantly better than the existing piledriver FX's, then maybe Haswell would be a lot less of a disappointment, for some people, in the enthusiasts market.
E.g. Solder, rather than TIM.

This lack of competition is a continuing argument in these forums, but no one really knows the answer. TBH intel has plenty of competition, but it is mainly not from AMD, it is from ARM and the entire ecosystem that surrounds it.

But when you consider the extensive research and validation that goes into producing a cpu, the cost of building foundaries, and precisely machining a chip with such small tolerences, can you really say that the cost is "SKY HIGH"? For the cost of a single night out on the town or the cost of a concert ticket you can buy a pentium cpu that has a tremendous performance except for the very most demanding users. And AMD is even cheaper, but I didnt bring that into the discussion because your comments seem to be directed only at intel.