AMD Shows Deneb (Phenom II) pictures of 6.2GHz CPUz etc

[informal]

500Ghz is easy as such...for a few transistors. But making 1 billion of them go at 500ghz is..a whole other story.

It's not the point(the sheer Ghz).The point is that ICs can work at such a low temps as IBM showed back in 2006,the thing you said was not possible.

edit: ...and he goes offline

[Shintai]

It's not the point(the sheer Ghz).The point is that ICs can work at such a low temps as IBM showed back in 2006,the thing you said was not possible.

edit: ...and he goes offline

Uhm..no? They didnt show a CPU running at -269 or -270C. No CMOS anywhere
Plus the the chip wasnt near absolute zero for sure. However they used the helium instead of LN2 etc as a more functional and controlled cooling (Plus helium is an inert gas unlike nitrogen). Its not just puring it down into a pot like with OC here. 350Ghz at room. 500Ghz cooled. But overdramatizing sells.

Here is also the actual article from NYtimes. And without the fake pic.
http://www.nytimes.com/2006/06/20/te...chip.html?_r=1

And enjoy this
http://arstechnica.com/news.ars/post/20060622-7117.html

And here is the actual pictures:
http://www.gatech.edu/upload/pr/tyf16878.jpg
http://www.gatech.edu/upload/pr/tzt16879.jpg

Its a 500Ghz radio transmitter basicly.

Ofcourse you could just use field effect transistors....

[clo007]

Yeah why don't they screen 100000 chips (probably next gen ones fused to look like current ones) and pick the best one. Then have their front man drive to Las Vegas and have an 'expert' overclocker run it on air with secret settings on a modified board and have an air conditioner blowing air through the cooler (hmm wonder why there was music playing in the background during the actual run) and video tape a single superpi run that hits one core and then claim it does 5GHz are cooled on 8 threads. Yup that sounds a better thing to do.

What do you all expect AMD to do from a part that hasn't officially launched yet? There are enough people who can be trusted that showed up at the event and saw everything with their own eyes. I believe that more than what I just described above.

wow...not bitter at all are you?

[Leeghoofd]

Think we are still talking CPU's here and that 500GHZ post is such a joke... why do some hold onto it ? think the person that googled it was also not aware of the real story behind that nice eyecatching title...the guy that chose that title must be a reporter of the Inq now or maybe of the Sun...

Give it all a rest... this thread is really getting nowhere with all this non stop posting about nothing... would be nice if a moderator cleaned up these threads and removed the zillions of useless posts... I have no clue what the point is of some posts : to show that you are more literate then another person, more technical skilled,...to show off how clever you seem to be ?

Thx V_rr for some early results... at least something usefull in this thread...

What do you all expect AMD to do from a part that hasn't officially launched yet? There are enough people who can be trusted that showed up at the event and saw everything with their own eyes. I believe that more than what I just described above.

Still shows me no proof about the discussion if the cpu's were also preselected or not (which so many seem to defend here that AMD did NOT preselect anything), if they had disabled certain stuff in the CPU to get it running faster etc... AMD seems to have something up their sleeve and I'm looking forward to it... BUT if it just is able to equal current socket 775 CPU's (Oc'ed ofcourse) then they remain a step behind and Intel will continue to dominate the OC'ing scene (I care zilch about servers, laptops etc...)

[AbelJemka]

BUT if it just is able to equal current socket 775 CPU's (Oc'ed ofcourse) then they remain a step behind and Intel will continue to dominate the OC'ing scene

Wrong if it can equal current socket 775 CPU but can OC higher it can be an interesting alternative on the OC'ing scene.

[generics_user]

Wrong if it can equal current socket 775 CPU but can OC higher it can be an interesting alternative on the OC'ing scene.

so true, i7 platforms cost 2 times more than amd ones

[Leeghoofd]

So you two both claim you would quit eg ya current Intel rig for a similar behaving AMD platform ? if we get 4 - 4.4Ghz (air and water) out of this Phenom II and at those speeds it eg just equals the current 775 line up I would be so dissapointed after all the hype we got...

My point is not many users will make the transition to an AMD platform if it just can keep up with the current socket 775 CPU's or barely beat it...

And why do you compare to I7 ? Lost me completley there

[Utnorris]

I think the point is that no one would make a lateral move to these chips, however if they were running an older chip and they now have the choice between AMD and Intel as we use to, they would go with AMD. Usually the costs associated with building an AMD system are less than an Intel system and not everyone is going to shell out between $700 to $900 to get into the I7's. It's a consideration when you are building a completely new system because yours is 2-3 years old, but if you just want to upgrade your 1 year old system then you will probably want to use as much of the old system, i.e. memory, that you can.

[Rammsteiner]

Oh emo

I think you miss the point of kinetic energy. Plus the high voltage that increases it. And then with the highly reduced flexibility at a colder temperature. It wont be slower. It just degrade very very fast as the transitors brake under pressure due to mechanical stress. Its basicly eletromigration and "hot electrons" at an extremely accelerated pace.

And before you say superconductor. They basicly have the completely opposite goal than CPUs in this matter.

Wut, Im no emo lol

Anyway, indeed Ive no 'n clue about kinetic energy involved in such situations, thanks for sort of clarifying. Although I already got the point that such low temps wouldnt do the CPU any good, how and what it would do...

So you two both claim you would quit eg ya current Intel rig for a similar behaving AMD platform ? if we get 4 - 4.4Ghz (air and water) out of this Phenom II and at those speeds it eg just equals the current 775 line up I would be so dissapointed after all the hype we got...

My point is not many users will make the transition to an AMD platform if it just can keep up with the current socket 775 CPU's or barely beat it...

And why do you compare to I7 ? Lost me completley there

But why do you ask that question since I only took one slight look at your rig to think that was a weird question coming from you, especially on this forum

Although I'd at least wait for AM3 boards to be out before making a jump from skt 775, AM2+ ain't dead yet either but as said, ain't making a lot of sense with AM3 being around the corner.

Also 4~4.4Ghz haven't been the noted average. 3.7~4.3Ghz was noted to be an eventual average. In the end Yorkfield ain't hitting 4~4.4Ghz on average either, if PhII would it would actually be a nice move (from tweaker pov). Also I dont see why 4~4.4Ghz would be a dissapointment in anyway I think that's actually very nice if true

[AbelJemka]

So you two both claim you would quit eg ya current Intel rig for a similar behaving AMD platform ? if we get 4 - 4.4Ghz (air and water) out of this Phenom II and at those speeds it eg just equals the current 775 line up I would be so dissapointed after all the hype we got...

My point is not many users will make the transition to an AMD platform if it just can keep up with the current socket 775 CPU's or barely beat it...

And why do you compare to I7 ? Lost me completley there

If PII and Penryn are comparable in terms of performance and if you don't already have a Core 2 Quad, PII940 may be much more appealling with unlocked multi than Core 2 Quad. Moreover getting 4GHz with a Core 2 Quad is not that easy even with a Q5500.

[Sparky]

On the 3 from 30 chips thing, it was 30 chips that had passed initial testing at the factory, then 3, and only 3, were tested and used for the overclocks...

So I will explain again, they had a bag of 30 processors that had passed the intitial testing at the factory after pouring, cutting, pinning and lidding the raw product, then they picked three ONLY at random from that bag, ran tests on those three alone, and used the best one.......

http://www.xtremesystems.org/forums/...0&postcount=84

[Lightman]

I was wondering why no one here noticed two key words coming from AMD demos:

- after first leaks of 6GHz+ run people attending it were saying about revision used for demo as RB-C2H (I can only guess H state High Performance process).
- from recent demo we have information that AMD showed AM3 DDR3 CPUs and also overclocked them to 4GHz+.

Now we have some space for speculation.
Is this mysterious RB-C2H revision simply a RB-C3 one which supposedly is reserved for AM3 parts, or maybe AMD demoed both revisions and no one pointed it clearly?
Will first retail AM2+ parts OC as well as demoed parts???? I think not as well, but well enough to get close to 4GHz on AIR at the beginning.
Was AMD running AM2+ and AM3 systems using only RB-C3 revision? For me it seems likely.

Anyone care to speculate more?? Opinions??
Anything is good for discussion if it's on TOPIC!!

[Frontl1ne]

On the 3 from 30 chips thing, it was 30 chips that had passed initial testing at the factory, then 3, and only 3, were tested and used for the overclocks...

And yes, the 30 chips may well have been cherry picked by AMD but keep in mind, they had a 66% chance of getting a PII that only clocks to 5.5GHz. How many people would have been accusing AMD of cherry picking the chip at the Austin, Texas demo if they only managed 5.5GHz at San Fransico! If AMD had in fact cherried the 30 chips, they would have ensured an average of close to 6GHz for all 30 chips to assure the skeptics that the 6.2GHz chip in Texas was not a cherry picked chip. But yet, they managed to pull out two 5.5GHz chips along with the 6GHz chip.

Will first retail AM2+ parts OC as well as demoed parts???? I think not as well, but well enough to get close to 4GHz on AIR at the beginning.

There was a CPU-Z validation of a PII that scored over 4GHz. It scored 4.1GHz on a SB600 motherboard. Link here

[Hornet331]

I was wondering why no one here noticed two key words coming from AMD demos:

- after first leaks of 6GHz+ run people attending it were saying about revision used for demo as RB-C2H (I can only guess H state High Performance process).
- from recent demo we have information that AMD showed AM3 DDR3 CPUs and also overclocked them to 4GHz+.

Now we have some space for speculation.
Is this mysterious RB-C2H revision simply a RB-C3 one which supposedly is reserved for AM3 parts, or maybe AMD demoed both revisions and no one pointed it clearly?
Will first retail AM2+ parts OC as well as demoed parts???? I think not as well, but well enough to get close to 4GHz on AIR at the beginning.
Was AMD running AM2+ and AM3 systems using only RB-C3 revision? For me it seems likely.

Anyone care to speculate more?? Opinions??
Anything is good for discussion if it's on TOPIC!!

Interesting observation, at least some guys keep there eyes open for the details.

But to bad its not possible to confirm anything, we need to wait for retail parts to see whats behind this.

[renzokuken]

6.2ghz!

*hopes this is possible on retail silicon*

[jonspd]

6.2ghz!

*hopes this is possible on retail silicon*

maybe 1 out of 1000 would do it but any more then that you can probably forget it.

[Donnie27]

Originally Posted by taurus_sel
Yeah why don't they screen 100000 chips (probably next gen ones fused to look like current ones) and pick the best one. Then have their front man drive to Las Vegas and have an 'expert' overclocker run it on air with secret settings on a modified board and have an air conditioner blowing air through the cooler (hmm wonder why there was music playing in the background during the actual run) and video tape a single superpi run that hits one core and then claim it does 5GHz are cooled on 8 threads. Yup that sounds a better thing to do.

What do you all expect AMD to do from a part that hasn't officially launched yet? There are enough people who can be trusted that showed up at the event and saw everything with their own eyes. I believe that more than what I just described above.

wow...not bitter at all are you?

Of course LOL! Even more of the Press, Webmasters and etc..... were at the First Conroe event. Yet, some of the same folks who say trust now, said Don't trust then. Same old double standards.

I believe the LN extreme overclocks because I know someone who at 8 only one of those reach 4GHz. The average is not 3.7 to 4.2GHz, they're say 3.4 to 3.8GHz (Bud's test 8 and one did 4GHz). Still a very nice improvement over Phenom 1. You'll not have to burn a IMC to get those speeds either :wink:

[Mav451]

I trust an XS member will release results when they (legally) can
On the AM3 vs. AM2+...I should probably pay attention, with the former requiring me to change RAM (damnit).

[Shadowmage]

I think you have the wrong idea of Intel ... we have full freedom to post what ever we want , as long as we do not break our NDA with Intel.
This is not the big brother you imagine ... Get use to it, we have our 1st Amendement right, or the 11th of the French constituion right ....

Thanks

Francois for President!

Callin you out on this. You're an exception as an Intel employee. Most employees don't have full freedom: they do it under the radar. Most employees do not have your 1st amendment rights.

How do I know? Let's say that the secret Intel employees who have browsed XS (not necessarily a member) have told me so. Why have they told me? It's because the same employees are actually upset because you're tarnishing Intel as a company with your lies about AMD. Don't criticize me: I'm just the messenger.

[Zucker2k]

Callin you out on this. You're an exception as an Intel employee. Most employees don't have full freedom: they do it under the radar. Most employees do not have your 1st amendment rights.

How do I know? Let's say that the secret Intel employees who have browsed XS (not necessarily a member) have told me so. Why have they told me? It's because the same employees are actually upset because you're tarnishing Intel as a company with your lies about AMD. Don't criticize me: I'm just the messenger.

This is unnecessary Why can't you just disagree without being so obnoxious? I use Intel products and I'm glad Francois is on this forum dropping little hints here and there about Intel products, if you don't like it, use the ignore button already!

[Shadowmage]

This is unnecessary Why can't you just disagree without being so obnoxious? I use Intel products and I'm glad Francois is on this forum dropping little hints here and there about Intel products, if you don't like it, use the ignore button already!

The problem is that he's doing it while being childish about AMD. Also I have some very reliable sources that state that there's a lot of things about Nehalem that he's not telling you. Ironically, there's more things "wrong" with the Intel Nehalem demo sponsored by DrWho than AMD's demo. If anything, he's giving Intel a bad name as far as publicity is concerned.

[YukonTrooper]

I hear the red, orange, and black coming.

[Shadowmage]

I hear the red, orange, and black coming.

If you mean the mods, don't worry, I've already checked with them.

[taurus_sel]

How do we know that? Let me remind you that AMD has been lying its ass off for the last two years, In fairness the last time they seemed bullish they hit the nail on the head, The hd4850 in my rig is testament to that, so they have a bit more credibility imho but everyone has to admit there has been some whoppers told this last two years and to just flat out believe that the 30 cpus they brought to a marketing event had no kind of selective binning is naive at best. Just for the record I would not believe any Intel marketing comment like this ether.

I am just speculating and could be wrong so lets come back to this thread when p2 is launched and feast on humble pie or show how magnanimous we can be

http://www.flickr.com/photos/amd_unp...ed/3047145934/

You want to know how we know that...look at this picture. Look beside Macci on the big screen tv. Yes, there are 4 cores enabled. And yes there is LN2 in the tube. And yes macci is doing the demo.

[Vapor]

I hear the red, orange, and black coming.

It's staying for now....let's not let it get worse though.