Intel to Retaliate to AMD Phenom II Overclocking Feat, Plans Demonstration at CES '09

[Miss Banana]

Don't start again. Fugger will get the mother of all unlocked chips. Why can't you all just be happy for him? Stop the ing.

I think as long as it happens in a respective way, it is okay to talk about why intel cherrypicks and unlocks CPU's as a response to AMD maybe doing the same.

On a sidenote, this whole event credits the importance of XS as a community, which is a cool thing I think.
What happens on XS matters!

[Charles Wirth]

Damn, I didnt see this thread

News to me, Im kinda busy that week too.

Cooper, Lots of good ones deleted. slow down on the damage control.

Im on my own schedule, it was just known publicly that Intel was coming out this time and media spinning this, hopefully AMD can confirm or does anyone have a link to where they got that info?

[NightCrawler™]

Uhm yeah.... I don't see any reason why my posts would have been deleted from that thread...

Anyways... I hope Fugger will let some I7's bleed.. but I still feel Intel taking this way too far..
It kind of reminds when AMD presented it's S939 CPU's...

[Zucker2k]

Damn, I didnt see this thread

News to me, Im kinda busy that week too.

Cooper, Lots of good ones deleted. slow down on the damage control.

Im on my own schedule, it was just known publicly that Intel was coming out this time and media spinning this, hopefully AMD can confirm or does anyone have a link to where they got that info?

Here guys, read the first 4 pages of this thread: http://www.xtremesystems.org/forums/...d.php?t=208374

DrWHO earlier indicated he was sure AMD had disabled some features; interestingly, Informal confirmed his suspicion? by actually confirming it (imho)

You gotta stop clutching at this straw soon.
The max clock AMD achieved with Phenom 65nm is 2.6Ghz ,which is 15% short of 3Ghz.Not that much.TLB issue made them dedicate the much needed time (for speed path optimiz.) for other things like fixing the damn bug in hardware.

Like i said,you need to stop living in the past.AMD's chip can work without failing in the range from -200 to +100 degrees Celsius,as Theo said here:
http://theovalich.wordpress.com/2008...re-i7-thunder/

Quoting Theo:

AMD decided that the time has come to regain the crown in overclocking, and the company means it. You can thank all of this to several key members of former ATI Technologies, now AMD GPG. I know it is not fair to name the few, since I can’t mention them all, but guys such as Godfrey Cheng, Ian McNaughton, Korhan Eben, Macci, and many unsung heroes combined with old-school AMD executives such as Patrick Moorhead, who keeps his desktop machines OC’ed to 3.2 GHz and arguing with IT Police… All in all, this effort brought a lot of fighting spirit to AMD and these guys are now biting at all sides. First with Radeon series, then with chipsets and now with CPUs.
Phenom II is going to be extremely overclockable. How overclockable? Much.
With good air-cooler, 4.0 GHz is a given on almost every Black Edition CPU that will hit the stores starting January 8, 2009. This is nothing special, since Intel can do the same with Core i7 series. But, with water-cooling or TEC-enhanced water-cooling (hint: CoolIT), you can easily reach 4.5 GHz and beyond.
But, the real show starts when we dip below the zero degrees Celsius/Centigrade. History of AMD and sub-zero cooling is quite interesting. AMD was the first manufacturer to showcase vapor chamber and liquid oxygen cooling (yes, oxygen, not nitrogen) with Athlon 550 MHz that reached 1.03 GHz and became the first CPU to pass the 1 GHz mark. As the time went by, AMD slowly started to move away from Swashbuckler attitude lead by Jerry Sanders and put a dull corporate Motorola face lead by Hector Jesus Ruiz. All of this is past now, since AMD put substantial effort to produce a CPU which operating temperature is massive 300 degrees Celsius!
Yes, you’ve read it correctly – from -200 to +100, you can be certain that your Phenom will not fail. On-die sensors are tweaked up and they will not lock the part at -100 or -20C, and you can use dry ice or LN2 to crank it up to the max.

The max reached so far by AMD team is… 6.0 GHz! Given that most LN2 clockers use dual-core CPUs to reach 5.5+ GHz speeds, seeing 6.0 GHz CPU running applications such as Crysis is a testament to monumental effort put by AMD’s CPU and Chipset team. Don’t think this was CPU team only effort – guys from Markham worked hard on Advanced Clock Calibration, or ACC. ACC is exactly “the secret sauce” AMD needed to unlock these levels of performance.
Hitting 4.0 GHz on air, 4.5 on water, 4.5+ on TEC+water, 5.6 GHz on dry ice and there was a post at 6.0 GHz. Given the fact that dry ice is “weaker” solution than LN2, it will be really interesting to see what will happen when hard-core overclockers get their hands on these 45nm ice-cold babies.
AMD is back. And yes, Intel’s Core i7 thunder has just been stolen completely. Core i7 is a great CPU, but quite expensive platform, with three channels of DDR3. AMD Phenom II comes out with 16 GB/s of memory bandwidth using only two channels, and clocking like there’s no tomorrow.
The best part of them all is the price: AMD Phenom II 940 Black Edition will set you back for 40% of the amount you have to shell out for Core i7 Extreme 965 and yet, it comes with a radically cheaper platform of equal or even better overclocking capabilities.

And DrWHO replies:

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

no more comments from me

So this is where the real debate lies, Theo claims every PII chip should do -200 to 100+, and Francois seems to claim that the tweaks were only applied to the bench-marked chips. Only time will tell.

On another note, Theo's comment about AMD's effort to claim the overclocking record should not be overlooked; imho it legitimizes every move Intel makes to counter than notion. So far, Intel is ahead, but like I said in another thread, they have learnt their lesson in the past and are not going to make the same mistake twice. I don't blame them if they're going after AMD, after all what was the purpose of AMD demoing the chip to the media, alonside a i965 XE?

It seems what AMD was hoping to avoid, a clockspeed war with Intel, is not going to happen. Personally, I love it; and I'm praying those PII tweaks come stock, who knows, Intel may follow suit and before you know it, we'll be getting BE (Blue Edition) Nehalems with unlocked multis (dreaming).

[cegras]

Yeah, my opinion on this is probably clear enough.

Still, I'm curious about the poopstorm that's going to down on Jan 9th (?).

[Macadamia]

January 9th.

LN2 galore. CES showfloor is snowing. AMD and Intel PR guys have snowfight and momentarily forget about competing in real life.

[TurboDiv]

...seeing 6.0 GHz CPU running applications such as Crysis is a testament to monumental effort put by AMD’s CPU and Chipset team.

IF Intel is going with this, i want to see the quote up there accomplished first by Intel and than the 7ghz fiasco. Let the fun begin...

[Jacky]

I think Intel doesn't care, Francoise Piednoels ego would be better. I have lost every respect i had for that man.

Yeah, damn it! It turns out he's an actual human with human feelings who likes to work for his company, how unfortunate.

I've only gained respect for him, when I first read his posts I thought he's a troll pretending to be someone from Intel.

Still I believe it's all FUD.

[Drwho?]

Damn, I didnt see this thread

News to me, Im kinda busy that week too.

Cooper, Lots of good ones deleted. slow down on the damage control.

Im on my own schedule, it was just known publicly that Intel was coming out this time and media spinning this, hopefully AMD can confirm or does anyone have a link to where they got that info?

Some "journalist" have a lot of imagination ... CES! hahahha
I was In Vegas, but that was yesterday ...

Accuracy of this news:

[flippin_waffles]

Or, and since it took so long for a reply, it's more likely that because of the backlash of this PR stunt, the event was pulled. Looks like not even 'the borg' could pull this one off.

[GAR]

I think there is no point in INTEL making a move, we all know intel overclocks good already.

[taurus_sel]

Some "journalist" have a lot of imagination ... CES! hahahha
I was In Vegas, but that was yesterday ...

Accuracy of this news:

Hey Dr. Knowitall, how do you know AMD is disabling sensors or doing anything special? Is this pure speculation or do you know something? Or is it not just the journalists that have a lot of imagination?

[taurus_sel]

Don't start again. Fugger will get the mother of all unlocked chips. Why can't you all just be happy for him? Stop the ing.

Yes Fugger will get the mother of all chips that no one else will get. How is that supposed to make anyone happy? How is that relevant? I'd be happy if I could buy something that overclocks at a decent price.

[Ghostbuster]

Yes Fugger will get the mother of all chips that no one else will get. How is that supposed to make anyone happy? How is that relevant? I'd be happy if I could buy something that overclocks at a decent price.

At least someone was able to lay hands on a "mother of all chips". I'm still waiting for that 3GHz Phenom which AMD demo'ed to appear somewhere.

[Drwho?]

Hey Dr. Knowitall, how do you know AMD is disabling sensors or doing anything special? Is this pure speculation or do you know something? Or is it not just the journalists that have a lot of imagination?

Well, Theo was pretty clean in his blog, no?
(I am insensitive to personal attack ...)

[taurus_sel]

At least someone was able to lay hands on a "mother of all chips". I'm still waiting for that 3GHz Phenom which AMD demo'ed to appear somewhere.

Under what rock have you been man. Do a search and you'll find all kinds of people cracking 3GHz on existing Phenoms.

[taurus_sel]

Well, Theo was pretty clean in his blog, no?
(I am insensitive to personal attack ...)

I don't read anything from Theo. Which blog are you referring to?

This is no personal attack, but you have to realize that you can't post up speculation as fact and hope no one calls you on it.

[Zucker2k]

IF Intel is going with this, i want to see the quote up there accomplished first by Intel and than the 7ghz fiasco. Let the fun begin...

And where did you get that idea; it's quite clear Theo was all over the place with that 6ghz figure; ie. OC or stable OC. Not to say this can't be done, but imho 6Ghz is a mighty oc on Nehalem, and if Fugger manages to get 7Ghz, 6ghz Crysis will be a walk in the park.

[Ghostbuster]

Under what rock have you been man. Do a search and you'll find all kinds of people cracking 3GHz on existing Phenoms.

I'm talking about that 3GHz on stock.

This demo http://www.youtube.com/watch?v=R7EZmYth6TM

[taurus_sel]

I'm talking about that 3GHz on stock.

This demo http://www.youtube.com/watch?v=R7EZmYth6TM

You have an excuse because you've been busy fighting ghosts. But man go here: http://www.anandtech.com/cpuchipsets...spx?i=3360&p=4 or just use google. The web is riddled with reports of 3+ GHz Phenoms.

[Drwho?]

I don't read anything from Theo. Which blog are you referring to?

This is no personal attack, but you have to realize that you can't post up speculation as fact and hope no one calls you on it.

hummmm ....

read better the post from theo:


"On-die sensors are tweaked up and they will not lock the part at -100 or -20C, and you can use dry ice or LN2 to crank it up to the max."

can you read it?

by the way, with Charles, we decided to stay without mods of the CPUs.

[Ghostbuster]

You have an excuse because you've been busy fighting ghosts. But man go here: http://www.anandtech.com/cpuchipsets...spx?i=3360&p=4 or just use google. The web is riddled with reports of 3+ GHz Phenoms.

That AnandTech was using the SB750 which did not exist when the 3GHz demo was shown.

I'm talking about cherry picked ones (like the one's FUGGER's getting). The 3GHz demo was using a standard heastsink. AMD didn't say it was overclocked either.

EDIT: Back then you should know the first Phenoms were the 9x00 series.

[Daveburt714]

Actually Theo Said:

"Yes, you’ve read it correctly – from -200 to +100, you can be certain that your Phenom will not fail. On-die sensors are tweaked up and they will not lock the part at -100 or -20C, and you can use dry ice or LN2 to crank it up to the max."

Not sure why you decided to leave out that part of the quote, but that would implie (at least to me) that it was a tweak made to the core. Not diasabled sensors...

[Zucker2k]

Actually Theo Said:

"Yes, you’ve read it correctly – from -200 to +100, you can be certain that your Phenom will not fail. On-die sensors are tweaked up and they will not lock the part at -100 or -20C, and you can use dry ice or LN2 to crank it up to the max."

Not sure why you decided to leave out that part of the quote, but that would implie (at least to me) that it was a tweak made to the core. Not diasabled sensors...

On die sensors control the thermal envelopes under which the cores operate. If they were/are "tweaked up/or down (as implied), then it means the cores will not shutoff at those tweaked up/down preset temps.

[taurus_sel]

hummmm ....

read better the post from theo:


"On-die sensors are tweaked up and they will not lock the part at -100 or -20C, and you can use dry ice or LN2 to crank it up to the max."

can you read it?

by the way, with Charles, we decided to stay without mods of the CPUs.

Yes I can read english and spell as well. And from your owns words:

"I am just going to be doing the same as our green buddies, dissable the thermal control, and let s see"

Again how do you know the green team is disabling anything? Clairvoyance? Guessing? Do you have the inside scoop?

Like I said, keep Theo out of this. I never read what he writes.