Intel to retaliate against AMD Phenom II OC

[BlueBlazer]

Originally posted by: GaiaHunter
Selling products at a loss to force smaller competition do die is considered illegal and governmental authorities go after it.

Unfortunately that doesn't seem to be the case, as Intel has been profitable always. So how does "selling products at loss" compute?

 

[GaiaHunter]

Originally posted by: BlueBlazer

Originally posted by: GaiaHunter
Selling products at a loss to force smaller competition do die is considered illegal and governmental authorities go after it.

Unfortunately that doesn't seem to be the case, as Intel has been profitable always. So how does "selling products at loss" compute?

Im talking to Nemesis about is 100$ and under cpus. And nowhere I specified a company.

 

[Kuzi]

Guys the Phenom II that reached 6GHz was cherry picked, so retail CPUs may not be as good. But I'm thinking many retail Denebs will reach 4GHz on air, because even the ES samples tested few months ago were reaching 3.6-4GHz on air.

As to IC7 reaching 5GHz on air, it was also hand picked. Don't forget Intel had a year to tweak their 45nm process, while AMD are on their first retail batch. I'm sure with newer steppings (AM3 versions?), Deneb will OC even higher than 4GHz on air.

 

[Nemesis 1]

Originally posted by: GaiaHunter
Selling products at a loss to force smaller competition do die is considered illegal and governmental authorities go after it.

Who said intel has to sell at a loss? Would it be legal for intel to cut prices to 10% margins . Still making Money. No intel can't do this by law. Its not right but thats the way the money changers wanted things. Its all about control. Intel still makes money at $100 apiece. Atom cost intel $6 to make. What you think those other Cpus cost intel to make.

 

[Nemesis 1]

Do you have a link to 6ghz P2??????? If so lets see it. With proof . Cpu-Z not a fake a real one. Also running superpi wouldn't hurt to see a run. If P2could do as claimed AMD marketing is from another world. On P1 they showed o/c s that never really occurred. Than they get super chip and offer no proof . Very strange proceedings.

 

[GaiaHunter]

Originally posted by: Nemesis 1

Originally posted by: GaiaHunter
Selling products at a loss to force smaller competition do die is considered illegal and governmental authorities go after it.

Who said intel has to sell at a loss? Would it be legal for intel to cut prices to 10% margins . Still making Money. No intel can't do this by law. Its not right but thats the way the money changers wanted things. Its all about control. Intel still makes money at $100 apiece. Atom cost intel $6 to make. What you think those other Cpus cost intel to make.

That is if you have a magical book where you can get the technology and the know how from.

Intel reduces profit margin, amd dies, US and EU jump on them and force them to split.

 

[piesquared]

Francois Piednoel not only stated that these were cherry picked chips, but that AMD were cheating. Let's just make sure that's on record.
Then goes on to cherrypick his own chip out of 100,000 processors to try to match it. Pretty amazing really, considering The Borg claimed 45nm was a no go without HKMG. It appears AMD's 45nm SOI w/ immersion litho is quite strong after all.

 

[Nemesis 1]

Yes it is . According to whats being said. Even tho P2 server has not shown any of this. Why is that? Well its intels Who saying AMD cheated. So will find that out in jan. I want to see all AMD P2 O/C to at least 4.0 ghz on air. AMds 45nm immersion appears by word of mouth to be very good. So far tho the one I am believing is Tony. He says its fact. I can't wait to see intels 32nm immersion . But than I want to see AMD 45nm. First. If its all that . Hay thats great.

Who is being alot more honest than you guys say. HE says YA I can cherry pick CPus also . 1 out of 100,000 not likely. Fugger and who are playing all . I had a pretty good setup that went back already. But the new system is just sick . You guys do what ever ya want. Fact is I have Money riding on AMD so please buy AMD . I will stay with the intel setup. Even tho I thought P1 was ok. But thats just me.

 

[Nemesis 1]

Originally posted by: GaiaHunter

Originally posted by: Nemesis 1

Originally posted by: GaiaHunter
Selling products at a loss to force smaller competition do die is considered illegal and governmental authorities go after it.

Who said intel has to sell at a loss? Would it be legal for intel to cut prices to 10% margins . Still making Money. No intel can't do this by law. Its not right but thats the way the money changers wanted things. Its all about control. Intel still makes money at $100 apiece. Atom cost intel $6 to make. What you think those other Cpus cost intel to make.

That is if you have a magical book where you can get the technology and the know how from.

Intel reduces profit margin, amd dies, US and EU jump on them and force them to split.

Actually I think Intel is thinking split up.. Something like AMD and the Moslems are doing. Intel could be alot stronger apart than as whole so long as its still has some say in the off shuts.

Intel SSD/ Intel chipsets/ Intel CPUs X86 fusion / AVX cpus fusion /. EPIC cpus./ Larrabee gpcpu. / Intel Fabs . I would like to see intel broken up . Tomorrow if it can be arranged. Counts number of holdings that could end up with . Ya kill this beast.

 

[eternalone]

I love this controversy it makes for good competition and prices and keep them both trying to develop better products we needed this to happen for AMD, its good for us consumers.

 

[Nemesis 1]

Wouldn't it be great if Intel did break up . And 2 differant parts design X86 cpus. That would get the monkey off their backs. Its a good possiability also . Think on that . Baby bells I mean intels. I like the sound of that.

If sandy bridge is infact considered an AVX cpu and not X86. That leaves AMD alone in the X86 cpu . A true monoply. Of course Intel will have x86 gpucpu.

I just wish i knew more about AVX vs X86. I just talked to brother inlaw yesterday . He asked what I thought of the new set up . I didn't say anything . He finely says is that a smile I am hereing . That's an affirmative. But I didn't think to ask him about AVX. Not sure I would get ans. at this time anyway.

 

[piesquared]

Originally posted by: Nemesis 1
Yes it is . According to whats being said. Even tho P2 server has not shown any of this. Why is that? Well its intels Who saying AMD cheated. So will find that out in jan. I want to see all AMD P2 O/C to at least 4.0 ghz on air. AMds 45nm immersion appears by word of mouth to be very good. So far tho the one I am believing is Tony. He says its fact. I can't wait to see intels 32nm immersion . But than I want to see AMD 45nm. First. If its all that . Hay thats great.

Who is being alot more honest than you guys say. HE says YA I can cherry pick CPus also . 1 out of 100,000 not likely. Fugger and who are playing all . I had a pretty good setup that went back already. But the new system is just sick . You guys do what ever ya want. Fact is I have Money riding on AMD so please buy AMD . I will stay with the intel setup. Even tho I thought P1 was ok. But thats just me.

I want to see all i7's O/C to 5.0 ghz on air, which Francois just claimed to have achieved. So by the same standards, you should expect all i7's to equal that mark.

 

[pm]

I want to see all i7's O/C to 5.0 ghz on air, which Francois just claimed to have achieved. So by the same standards, you should expect all i7's to equal that mark.

CPU speeds are on a normal distribution. http://en.wikipedia.org/wiki/Normal_distribution There will some fast ones and there will be some slow ones, and then there will be a lot in the middle. If you go far enough down the curve, you can find some really slow parts.

Why is that? Well its intels Who saying AMD cheated.

I read every thread that I could find over at XS on this, and I never saw anyone from Intel saying anyone from AMD cheated.. if anyone finds a post like this, I would like to see a link. Francios said that it's possible to cheat - not that that this occurred.

While he could (should) have chosen his wording a bit better, he has a point about not trusting any manufacturer overclocking. It's one thing for some guy to tweak his system, but when manufacturers do it, it's best to take it with some level of distrust. And this is any manufacturer - with CPU's and GPU's or any other synchronous complex CMOS integrated circuit, there's a lot of behind-the-scenes tricks that can be done. Some are common-sense (disable overcurrent checks, disable thermal limits), some are semi-legit (pick fast parts, use LN), and some are very dubious (hypothetically, disable the L2 cache because you have a bunch of speed paths reading the L2, and then choose benchmarks that don't hit on the L2 at all, turn-on single-issue mode and don't run any benchmarks... the list is extensive including really funky ones like re-write the microcode or using arch-break handlers... I could go on).

Francois could have chosen his wording better, but I agree with his sentiment - which is not to put a lot of faith in any manufacturer demo'd overclocks as indicative of anything except that manufacturers are very good at tweaking their parts. This probably applies to more than just CPU's and GPU's too... cars, planes, anything where manufacturers are trying to show speed.



* Not speaking as a representative of Intel Corp. *

 

[Keysplayr]

Originally posted by: Nemesis 1
Actually your wrong . If I am not a stock holder of record . Your right. But If I can show I bought Stock after the news release . Ya As A stock owner of record I can indeed sue. Because the news was hype. You can't do that. Not to stock holders. To the public ya. You can lie till your blue in the face to those.

So you would sue even with advanced knowledge of what really happened. You'd try to convince a jury that you were suckered in by the 6GHz PhenomII. We'll, let me know where the trial is, and I'll present this thread to them.

It's beyond ridiculous now.

 

[bradley]

Originally posted by: piesquared
Francois Piednoel not only stated that these were cherry picked chips, but that AMD were cheating. Let's just make sure that's on record.
Then goes on to cherrypick his own chip out of 100,000 processors to try to match it. Pretty amazing really, considering The Borg claimed 45nm was a no go without HKMG. It appears AMD's 45nm SOI w/ immersion litho is quite strong after all.

I've enjoyed reading Francois messages over the years, well before his proverbial coming out of the closet. I'm disappointed with some of his recent statements. I believe they show a slight lack of character and discipline.

Though professionalism is in short supply on today's tech forums, and maybe it's slightly contagious. Francois has also become more marketer now than engineer. To be honest, it's pretty silly to make heroes out of people with dry ice and tubes. Where's the ingenuity there?

I'm also unsure of just how much marketing pressure is coming from his employer. However, it should speak great volumes about AMD's latest to see this type of clamor from Intel. They definitely see AMD as greater threat than usual, which can only be good for consumers. I just want to see fairer comparisons between the two products and then let the buyer decide. Although we're probably not going to see average i7 chips clocking to 5GHz on air for quite some time, which makes Dr. Who's exertion fairly moot.

 

[Idontcare]

Originally posted by: bradley
To be honest, it's pretty silly to make heroes out of people with dry ice and tubes. Where's the ingenuity there?

I'm sure you meant to say "where's the practicality" and not "where's the ingenuity".

I spent 5 yrs working with cryogenic systems (Dry ice, LN2 and Liquid Helium among others) and to say these folks lack ingenuity in accomplishing what they accomplish is just ignorance. It's not as simple or straightforward as those little snippets you read on websites or the handful of photos you might look at on the web. For every success that you get to read about there is a bin full of trashed hardware and empty dewars from LN2 in the corner not to mention the frostbit fingers.

But it is a valid statement to say they shouldn't be made into hero's simply for the sake of their epic overclocks, unless doing that has saved a kitten or baby jesus or starving children in Somalia. Nothing in this industry is so demanding or challenging as to warrant saintdom upon completion of the journey.

And it is equally valid to say that despite all the ingenuity that does go into achieving 6GHz OC's with LN2 it just isn't practical experience or an achievement that is any help for the common enthusiast.

 

[piesquared]

Originally posted by: pm


...I read every thread that I could find over at XS on this, and I never saw anyone from Intel saying anyone from AMD cheated.. if anyone finds a post like this, I would like to see a link. Francios said that it's possible to cheat - not that that this occurred...


* Not speaking as a representative of Intel Corp. *

While you are right that he was very careful not to explicitly state that AMD cheated, he absolutely implied it.

http://www.xtremesystems.org/f...p=3444648&postcount=68

Quote:

Originally Posted by informal
All that is fine and dandy,but you're forgetting one very important thing:
AMD has no room for that kind of stunts! They would loose A LOT if this is the case.
And honestly i doubt that this demo was a"stunt".AMD has nothing to gain since it would hurt the very much if people fail to reach close to teh claimed clocks.Remember: air cooling =3.9-~4GHz ,wc=4+Ghz,dry ice =5+Ghz,LN2=6+Ghz.
It's simple as that.

PS Intel doing the Ln2 demo after core i7 was long released and just after Deneb LN2 sessions would look a LOT like knee-jerk reaction...

Quote:
AMD has tweaked the on-die sensor to not lock the part

no more comments from me

http://www.xtremesystems.org/f...p=3444580&postcount=61

This is why at the end, what matter is to have a curve with at least 4 frequencies to make sure that nothing was defeatured to get to the high clock (When it is manufacturer demo)

For the moment, I will pull a side those tricks and see what I can get with LN2, and to make sure it is true, I ll do it with a famous OC master , then, I ll pull my tricks ...

And then there's this:

http://www.xtremesystems.org/f...=3449302&postcount=278

Dudesss, I am like anybody else, I like competition, this is fun when AMD start showing the teeth

Before I compete, I like setting up the rules, and explain the themes ... I just did that. Some people think that I was "shooting" at AMD, but in fact. I was just spelling the rules of the game (Since they did dissable the thermal diode, allowing them to do the trick I did to get to 4.8Ghz)

Guys, I really want to be fair and honnest, do not give me the bad guy shirt here, I went open 100%.

Francois

You can find them all here:

http://www.xtremesystems.org/f...=10314622&pp=25&page=2

We'll see if he's right.

 

[Idontcare]

Originally posted by: piesquared
We'll see if he's right.

And then what? Right or wrong, why does any one care?

The passion in these threads (here and on XS) just baffles me, it's like we warped over the P&N forum for old times sake.

 

[thilanliyan]

Originally posted by: pm
I read every thread that I could find over at XS on this, and I never saw anyone from Intel saying anyone from AMD cheated.. if anyone finds a post like this, I would like to see a link. Francios said that it's possible to cheat - not that that this occurred.

I think he implied that AMD turned off some temp sensors.

EDIT: Piesquared beat me to it.

 

[pm]

Originally posted by: thilan29

Originally posted by: pm
I read every thread that I could find over at XS on this, and I never saw anyone from Intel saying anyone from AMD cheated.. if anyone finds a post like this, I would like to see a link. Francios said that it's possible to cheat - not that that this occurred.

I think he implied that AMD turned off some temp sensors.

EDIT: Piesquared beat me to it.

Thanks for the links. I know it took a while to look it up and I appreciate it. I was looking to see if there was anything in there that I personally thought was out of line, but - while I think he could have been more tactful... including not thread crapping - I didn't see anything over the top.

As far as temperature sensors, I don't see how that's 'cheating"... It seems to me that if you want to overclock in extreme temperature regions, you have do this.

 

[dmens]

Originally posted by: pm
While he could (should) have chosen his wording a bit better, he has a point about not trusting any manufacturer overclocking. It's one thing for some guy to tweak his system, but when manufacturers do it, it's best to take it with some level of distrust. And this is any manufacturer - with CPU's and GPU's or any other synchronous complex CMOS integrated circuit, there's a lot of behind-the-scenes tricks that can be done. Some are common-sense (disable overcurrent checks, disable thermal limits), some are semi-legit (pick fast parts, use LN), and some are very dubious (hypothetically, disable the L2 cache because you have a bunch of speed paths reading the L2, and then choose benchmarks that don't hit on the L2 at all, turn-on single-issue mode and don't run any benchmarks... the list is extensive including really funky ones like re-write the microcode or using arch-break handlers... I could go on).

agreed. i caught some flak in another thread for saying this is a desperation move by AMD. by doing such a demo, the manufacturer could be implying any customer can achieve the same result with an off-the-shelf part. such a demo would use a part with an in-die variation measurement so far off the average it probably wouldn't even make it to market because it is far off the maximum specified TDP under any circumstance.

even if AMD did not cheat besides custom fusing to disable to thermal/current limits, the implied possibility of a customer duplicating this result with a market part is misleading on its own.

that said, if intel did the same thing, i'd make the same claim. that goes for the old northwood/prescott demos.

 

[thilanliyan]

Originally posted by: pm
Thanks for the links. I know it took a while to look it up and I appreciate it. I was looking to see if there was anything in there that I personally thought was out of line, but - while I think he could have been more tactful... including not thread crapping - I didn't see anything over the top.

As far as temperature sensors, I don't see how that's 'cheating"... It seems to me that if you want to overclock in extreme temperature regions, you have do this.

Thanks go to Piesquared for the links.

About the temp sensor, yeah I figured doing something like that would be necessary to get very low temps and I'm guessing anyone going for extreme OCs on any Intel chip would have to do the same wouldn't they?

 

[piesquared]

Ok, now let's steer this back in the right direction. It seems to be taking a sudden shift to the left. Man, at the first opportunity.....

First of all, where is the proof that ANY temp sensors were disabled? Besides what Francois is alledging, which is the whole point of my post. This is exactly how this FUD spreads, so thanks for legitimizing my point. Francois comes along, makes an allegation, and it becomes gospel. Again, there is no proof that ANY temp sensors were disabled.

His assumptions are based on what i7 can, or cannot do. I'd like to know were the temp sensors disabled on Core2 with all those results on nitrogen? Because I don't remember this controversy then.

 

[dmens]

Originally posted by: piesquared
First of all, where is the proof that ANY temp sensors were disabled?

at 1.9V core? i can guarantee you any kind of throttling has been disabled. dynamic power scales squared with voltage, static power scales linearly.

 

[piesquared]

Assuming cpuz is anywhere near accurate of course.