Intel Explains 45nm Delays, Errata

[Monkeywoman]

More than a few people noticed Intel's roadmap originally slated 45nm Penryn desktop quad-core processors for January, only to have the company change the hard launch date to a not-so-firm "Q1 2008." So what happened? In a series of interviews with Intel, the tale of quad-core Penryn began to unfold.

Intel engineers, speaking on background, detailed the problem. "Intel is very sensitive to mean time to failures. During a simulation, at high clock frequencies, we noticed an increase of potential failures after a designated amount of time."

He continues, "This is not acceptable for our customers that require longterm stability. It's a showstopper."

Previous reports of errata degrading the L2 and L3 cache performance were described as "false." Microcode and BIOS updates issued by Intel since November do not fix or address the "showstopper" bug affecting the launch of the quad-core Q9300, Q9450 and Q9550 processors.

The condition does not affect Xeon quad-core processors. Intel tells DailyTech this is because Xeon uses a different stepping than the quad-core processors, which fixes this simulated condition. The quad-core 45nm Extreme Edition processor launched in November is also unaffected.

The company would not detail when the processors, originally scheduled for a January 20 launch but announced at CES last week, will see the light of day. Conservative estimates from ASUS and Gigabyte put the re-launch sometime in February. Intel completely removed its January 20 launch from its December 2007 roadmap and has not issued a new roadmap since.

Intel spokesman Dan Snyder says more. "We publicly claimed we will launch its 45nm mainstream processors in Q1 2008, and that's exactly what we did." In fact, the company announced 16 new 45nm processors last week; most of which already shipped to manufacturers -- with the exception of the quad-core desktop variants affected by the showstopper simulation bug.

Taiwanese media was quick to pin the simulated problem on complacency and lack of competition from AMD. Intel employees quickly denied the allegation, with the additional claim that the report was "humorous."

At CES last week, Snyder elaborates. "The tick-tock model prevents Intel from missing its launch dates. If the "tock" team misses a target date, it doesn't affect the "tick" team."

Tick-tock, the strategy of alternating cycles of architecture change and process shrink, became official company policy on January 1, 2006.

As to why the new Macbook Airs still use the 65nm Core 2 Duo processors? Even after Foxconn alluded the new notebooks would get 45nm treatment? Another Intel spokesman declined to respond, only stating, "Our partners are free to choose any of Intel's currently supported processors." Anand Shimpi explores this more.

Source: http://www.dailytech.com/Intel+Explains+45nm+Delays+Errata/article10362.htm

[Eastcoasthandle]

Will this turn out to be 3+ pages as found in AMD's errata issues?
Glad they caught it before it hit market...kinda cut it close but at least it's caught.

[Monkeywoman]

they launch too fast and AMD will die and they will be under the monopoly act. Intel is making it slow and painful

[Zytek_Fan]

Intel is doing this to give AMD time. They know their ass is on the line when it comes to product launches now while AMD is down.

[Planet]

Well if the bug is there why as 9650 released?

[ThatGuy16]

Didn't we hear of some people with having their ES chips die on them while overclocking? Seems like i was reading this somewere.. :shrug:

Atleast they caught the problems before release. :up:

[Zytek_Fan]

Didn't we hear of some people with having their ES chips die on them while overclocking? Seems like i was reading this somewere.. :shrug:

Atleast they caught the problems before release. :up:

QX9650's are really sensitive to voltage changes

[informal]

AMD has nothing to do with these delays.Anyone who thinks this way needs a reality check asap.

[Brother Esau]

Can anybody say Intel Frat House:rofl:

Intel has fault to you know?

[MpG]

So they didn't have this bug on the dual-cores?

[Darkmind]

Gah, my new water cooled water build is gonna have to spend another month or more on the work bench.:shakes:

[Donnie27]

Can anybody say Intel Frat House:rofl:

Intel has fault to you know?

True. They're lying. Since when is lying good? It is Intel's fault for lying. Lying = Bad. Sorry, that's not being a Fan, or is it? We believed it was a Bug on the 1133MHz at the time because it represented Intel best when the needed it badly. Mobo-MTH Bugs left them High and Dry with little or nothing. SO please understand when the very convenient Bug that only affect targeted products Intel doesn't need to rush out anyway leaves folks a litte skeptical.

[WeStSiDePLaYa]

they launch too fast and AMD will die and they will be under the monopoly act. Intel is making it slow and painful


Yeah, Intel planned to bug their cores to give AMD time.....:rofl:

[Cold Fussion]

Well if the bug is there why as 9650 released?

Did you read the first post? :rolleyes:

[fragmasterMax]

basically how i interpreted the first two paragraphs was that the penryn quad will be extremely stable at very high clock speeds, i cannot wait to obtain one.

[jinu117]

If anyone remembers the fiasco with intel P90...
This won't surprise anyone to be honest. In fact, there were few more after that was like p90 but that kind of went quiet as it was even less likely to be utilized. Some times unforseen things can happen and this might be one of them. Better to wait to get reputalbe source of information than start accusing right away. Intel to recall CPU in massive number can be.... COSTLY.

[savantu]

Intel managed to cause FSB integrity to go outside desired limits in laboratory controlled testing on cheap mobos.As a result , instead of asking mobo manufacturers to have tighter design rules , they do that on Yorkfield CPUs.

Xeon and Xtreme Editions CPUs aren't affected because they use steppings which have undergone more stringent design rules and extended validation.

[zakelwe]

ie they are stopping people with less money overclock as well :D

That will teach them

Regards

Andy

[K404]

Intel engineers, speaking on background, detailed the problem. "Intel is very sensitive to mean time to failures. During a simulation, at high clock frequencies, we noticed an increase of potential failures after a designated amount of time."

He continues, "This is not acceptable for our customers that require longterm stability. It's a showstopper."

Previous reports of errata degrading the L2 and L3 cache performance were described as "false." Microcode and BIOS updates issued by Intel since November do not fix or address the "showstopper" bug affecting the launch of the quad-core Q9300, Q9450 and Q9550 processors.

1/ The models delayed run at slower Freq. than the released QX9650, so that makes no sense.

2/ Intel dont have L3 cache.

WTF?

[Pete]

The bug is AMD

[Shintai]

1/ The models delayed run at slower Freq. than the released QX9650, so that makes no sense.

2/ Intel dont have L3 cache.

WTF?

I think they dont see QX9650 as an issue. Since in short its a 4layer vs 6layer PCB issue. So to use a 45nm quad on some cheap motherboard, then you can have issue. Same reason no servers are affected since they are 6+ layers.

And ye..L3 cache...WTF!

[adamsleath]

The quad-core 45nm Extreme Edition processor launched in November is also unaffected.
cheaper quads dont work
http://tbn0.google.com/images?q=tbn:S7FiIVsTHx1snM:http://inverse.physics.berkeley.edu/archives/nelson.gif

[quadnad]

AMD has nothing to do with these delays.Anyone who thinks this way needs a reality check asap.


The bug is AMD

So nobody thinks this is a great attempt at milking their current lineup? AMD hasn't released anything that needs a response from Intel, and so they've delayed the launch of their new processors in order to milk what they have going. It isn't terribly surprising, especially now that the revised Phenoms from AMD have been delayed.

[Iconyu]

I also think this is Intel seeing a easy method of offloading it's 65nm backlog at full price, demand has actually gone up on the Q6600.

[AMDDeathstar]

There is part of the sentence missing from the orginal article about L3 cache


Previous reports of errata degrading the L2 and L3 cache performance were described as "false" -- desktop Penryn processors do not even have L3 cache. Microcode and BIOS updates issued by Intel since November do not fix or address the "showstopper" bug affecting the launch of the quad-core Q9300, Q9450 and Q9550 processors.

[Donnie27]

I also think this is Intel seeing a easy method of offloading it's 65nm backlog at full price, demand has actually gone up on the Q6600.

QFT IMHO!

[quadnad]

There is part of the sentence missing from the orginal article about L3 cache


QFT IMHO!

tell it like it is! *snap snap*

[Donnie27]

tell it like it is! *snap snap*

See the thread about Intel price cuts? Look for a hint?

[quadnad]

See the thread about Intel price cuts? Look for a hint?

I completely agree with the both of you. Intel doesn't have a real reason to release this refresh, not with AMD standing still.

[Donnie27]

I completely agree with the both of you. Intel doesn't have a real reason to release this refresh, not with AMD standing still.

You got it! It seems some folks still can't admit AMD is standing still. Simple competition or the lack there of.

[Blasar]

I completely agree with the both of you. Intel doesn't have a real reason to release this refresh, not with AMD standing still.

Was with that kind of moves that AMD got F:bananal: :bananal: :bananal: :bananal: :bananal: :bananal: by INTEL when they released the Core Architecture, 1 year and a half ago.

[Ace123]

for the sake of competition, we can only PRAY that amd will try the same thing again. Nobody stands still in a race, unless they are planning a trick.

[quadnad]

for the sake of competition, we can only PRAY that amd will try the same thing again. Nobody stands still in a race, unless they are planning a trick.

Oh no, AMD isn't actually standing still; I just mean that they're doing so badly atm and for the foreseeable future that they may as well be standing still.

Look, I'm an AMD fanboy. But I'm not one to overlook facts, or to ignore the truth (I'm a geek afterall, numbers speak volumes to me). Thats why a Q9450 will be holing up in my rig as soon as they're available.

[FZ1]

AMD has nothing to do with these delays.Anyone who thinks this way needs a reality check asap.

You are mostly correct. If you had a product that needed revisions before it could be released and had no real competition, would you rush it out the door at an extra expense or take your time and get it right? So, there is some indirect influence on the actual release date but the "delay" would have occurred regardless of AMD.

[BrowncoatGR]

This thread is funny. Of course Intel is announcing imaginary bugs for their cpus, why wouldnt they... :p

[Skratch]

AMD has nothing to do with these delays.Anyone who thinks this way needs a reality check asap.

Its all about milking as much money as possible bud,there is no comp from AMD and intel only put out its 1400 dollar chips to milk us dry.There is no reason at all to release a new quad core when they already have the market with there q6600 ect...If AMD had a decent chip out now you would of sceen intel go full force with there 45nm chips.


They want to sell off all of its 65nm quads and hold out as long as they can being on top and then hit hard when AMD has something decent.

[naokaji]

There is no reason at all to release a new quad core when they already have the market with there q6600 ect...

not entirely true, because they dont want people to use q6600 forever, they not only want everyone who doesnt have a intel cpu to buy one, but they also want to sell those that allready have intel cpus some new ones.

[Donnie27]

for the sake of competition, we can only PRAY that amd will try the same thing again. Nobody stands still in a race, unless they are planning a trick.

QFT!

[Donnie27]

not entirely true, because they dont want people to use q6600 forever, they not only want everyone who doesnt have a intel cpu to buy one, but they also want to sell those that allready have intel cpus some new ones.

Intel knows it can get what it wants for Penryn. They also know they have to cut prices on most of the older Quads. This their is way to charge more for those old ones for a longer period of time. Meanwhile, stockpiles of the new ones are built up. Once the market is flooded with these most folks will quickly forget Qxx00 whatever LOL! There's really no reason for Intel to make that happen right now or even in a month from now.

@BrowncoatGR, Intel as always reported errata why doubt they wouldn't? It's not like they've never done it before:rofl:

[WeStSiDePLaYa]

Do you people realize that even though they have the lead, they still have to make deadlines to keep investors happy. If they miss deadlines, even when ahead, investors will be pissed.

Your delusional if you think they are making this up. It is also considered FRAUD to do that. And Intel wouldn't risk it when it already has several agencies sticking their head into Intel's ass looking for :banana::banana::banana::banana: to hit them with.

EDIT: Also, why would the purposely delay the 45nm tech? They can fit 30% more cores per wafer than with their 65nm tech. That means 30% more chips per wafer.

ALSO, Intel's profit per chip is very low right now compared to what it used to be, because of the price war AMD is giving them. 45nm tech will help them regain profit margins.

[FZ1]

Do you people realize that even though they have the lead, they still have to make deadlines to keep investors happy. If they miss deadlines, even when ahead, investors will be pissed.

Your delusional if you think they are making this up. It is also considered FRAUD to do that. And Intel wouldn't risk it when it already has several agencies sticking their head into Intel's ass looking for :banana::banana::banana::banana: to hit them with.

EDIT: Also, why would the purposely delay the 45nm tech? They can fit 30% more cores per wafer than with their 65nm tech. That means 30% more chips per wafer.

ALSO, Intel's profit per chip is very low right now compared to what it used to be, because of the price war AMD is giving them. 45nm tech will help them regain profit margins.
QFT!:up:

[Epsilon84]

I don't think AMD has any direct influence on THIS particular delay, but their lack of competitiveness has certainly harmed the industry as a whole.

I'm not sure if people realise this but since C2D was released Intel has pretty much sat on it's ass the whole time. Sure, there have been price cuts but technologically, what has Intel done in the past 18 months apart from 45nm? C2D launched at 2.93GHz in July 2006, we are still stuck at 3GHz today...

I find it hard to believe if there was actual competition from AMD that Intel would be sitting at 3.0GHz with their feet up on the couch!

[Shintai]

I don't think AMD has any direct influence on THIS particular delay, but their lack of competitiveness has certainly harmed the industry as a whole.

I'm not sure if people realise this but since C2D was released Intel has pretty much sat on it's ass the whole time. Sure, there have been price cuts but technologically, what has Intel done in the past 18 months apart from 45nm? C2D launched at 2.93GHz in July 2006, we are still stuck at 3GHz today...

I find it hard to believe if there was actual competition from AMD that Intel would be sitting at 3.0GHz with their feet up on the couch!

Arhh...thats abit too simplifying. We had a 2.93Ghz DUALCORE in july 2006. And we have a 3.0Ghz QUADCORE now and have had for a while. So a year to year we still have over a 100% performance increase.

Even with AMD pushing hard we wouldnt be much above. We get 3.2Ghz now basicly. We could get maybe 3.4-3.6Ghz then. But at a cost of TDP.

[Epsilon84]

Arhh...thats abit too simplifying. We had a 2.93Ghz DUALCORE in july 2006. And we have a 3.0Ghz QUADCORE now and have had for a while. So a year to year we still have over a 100% performance increase.

Even with AMD pushing hard we wouldnt be much above. We get 3.2Ghz now basicly. We could get maybe 3.4-3.6Ghz then. But at a cost of TDP.

OK, let's talk Kentsfield then. Introduced at 2.66GHz in November 2006, we have gained 333MHz in the space of 14 months!

I know as overclockers we haven't really suffered since we can crank up the CPUs to where they SHOULD be :up: but you can't really argue that if Intel had their backs to the wall that they would still be at 3GHz today! :shakes:

Intel plans to release nothing faster than 3.2GHz for the remainder of 2008, which is ridiculous considering it only consumes 66W according to Anandtech. http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3184&p=3

Intel could easily release a 3.8GHz quad TODAY and remain under 130W TDP.

[adamsleath]

dont tell me intel are going to sit on their hands just because AMD with its pitiful marketshare doesnt "compete" :rolleyes:
if intel continue to make profit with current lineup y would they give a rat's turd?

but it's always in their best interest financially to release refreshes / new 'products' - it's just a matter of time.
just to state the bleeding obvious.

they need / want to release 45nm quads to perfect their new 45nm process dont they?

prior to nehalem.

or maybe they could can the whole 9xxx lower order series - and still make good profits.

[Shintai]

OK, let's talk Kentsfield then. Introduced at 2.66GHz in November 2006, we have gained 333MHz in the space of 14 months!

I know as overclockers we haven't really suffered since we can crank up the CPUs to where they SHOULD be :up: but you can't really argue that if Intel had their backs to the wall that they would still be at 3GHz today! :shakes:

So you got a quad and it pushed what we can do on 65nm. Seriously, what do you expect? If you havent noticed x86 is at a dead end. Only die shrinks enable us to push further really.

Tell me, between 2000 and 2008 when we had a major speedbost? Besides going from single to dual and dual to quads. Core->Core 2 is about 20%. Athlon to Phenom is about 10%?

If you havent noticed, it got SQUAT with competition. But rather architectual design limits. Hell, look at Nehalem. 10-25% faster than Penryn approx. And thats with an IMC.

Any boost we have today comes with a process shrink, or is the benefit of the shrink. Aka more cores and/or faster speed.

[Shintai]

dont tell me intel are going to sit on their hands just because AMD with its pitiful marketshare doesnt "compete" :rolleyes:
if intel continue to make profit with current lineup y would they give a rat's turd?

but it's always in their best interest financially to release refreshes / new 'products' - it's just a matter of time.
just to state the bleeding obvious.

they need / want to release 45nm quads to perfect their new 45nm process dont they?

prior to nehalem.

or maybe they could can the whole 9xxx lower order series - and still make good profits.

Intel also needs to compete with itself. Without innovation and new speed. Why buy a new CPU?

And Intel really really want 45nm out and in full gear. They get better speedbins and they save some 30% in production cost per CPU. 143mm2 vs 107mm2 on dualcores.

[adamsleath]

22nm is planned in the not so distant future..which is half again 45nm :eek::lol:
the sky's the limit.

[Epsilon84]

So you got a quad and it pushed what we can do on 65nm. Seriously, what do you expect? If you havent noticed x86 is at a dead end. Only die shrinks enable us to push further really.

Tell me, between 2000 and 2008 when we had a major speedbost? Besides going from single to dual and dual to quads. Core->Core 2 is about 20%. Athlon to Phenom is about 10%?

If you havent noticed, it got SQUAT with competition. But rather architectual design limits. Hell, look at Nehalem. 10-25% faster than Penryn approx. And thats with an IMC.

I am merely saying that Intel isn't pushing nearly as hard as they would be if they were challenged.

Between 2000 to 2008, are you kidding? We were running 1GHz P3s and Athlons in 2000. By 2003 we had 3GHz P4s and 2GHz Athlon XPs.
While the clockspeed of Core2 hasn't increased over P4, it has doubled the IPC of Netburst, so effectively a 3GHz Core2 is equal to a 6GHz P4, and I'm not even taking into account the extra core.

Competition drives innovation, and if Intel isn't challenged, there is less incentive to innovate. Just look at nVidia as a classic example, they have had the 8800GTX for nearly 18 months, because of lack of competition from ATI/AMD they have dramatically slowed down their development cycle.

[Shintai]

22nm is planned in the not so distant future..which is half again 45nm :eek::lol:
the sky's the limit.

22nm is perhaps one of the last shrinks. We are close to have reached physical limits. 16nm might be possible too. But it sure wont be with a 2 year cycle. A 22nm factory might cost around 12-15billion$. If you notice more and more companies also gets thrown of the wagon. Because the costs gets too high.

16nm is expected to be reached around 2018. (Or 2013-2014 by Intel if they can keep the tick tock.)

[smee]

So the quad-core Q9300, Q9450 and Q9550 processors are going to be delayed. But the the dual-cores are still coming on the 20th right??

[Donnie27]

So the quad-core Q9300, Q9450 and Q9550 processors are going to be delayed. But the the dual-cores are still coming on the 20th right??

Right!

[Donnie27]

I am merely saying that Intel isn't pushing nearly as hard as they would be if they were challenged.

Between 2000 to 2008, are you kidding? We were running 1GHz P3s and Athlons in 2000. By 2003 we had 3GHz P4s and 2GHz Athlon XPs.
While the clockspeed of Core2 hasn't increased over P4, it has doubled the IPC of Netburst, so effectively a 3GHz Core2 is equal to a 6GHz P4, and I'm not even taking into account the extra core.

Competition drives innovation, and if Intel isn't challenged, there is less incentive to innovate. Just look at nVidia as a classic example, they have had the 8800GTX for nearly 18 months, because of lack of competition from ATI/AMD they have dramatically slowed down their development cycle.

QFT! Even for folks throwing the "Your delusional if you" line around:rofl: sheesh! Why do some folks seem so angry must be too much Ganstaplaya Rap Music:D

@shintai Intel can make just as much money from a Mature Process than a smaller process. All of the costs and or overhead associated with 65nm have long been paid down and Paul Otellini himself said yields were through the roof. They don't have to go at breakneck speed when they can pass the competition while at a slow walk.

In your honest opinion do you think Intel would be doing this if Phenom was 15 to 25% faster depending on the app? IMHO, Intel would be selling it and warning folks not to use it on the cheap 4 layer boards. There is a precedent for this. Look at Conroe on i975? A very similar thing happened with VRM and etc...

[smee]

Right!

Sweet!!!!!!
That means all the current procs are coming down in price :)

Thanks Don!!

[Donnie27]

Sweet!!!!!!
That means all the current procs are coming down in price :)

Thanks Don!!

Yes they will from now to April 20th.

[fragmasterMax]

Sweet!!!!!!
That means all the current procs are coming down in price :)

Thanks Don!!

Not necessarily. The fx60 (it's price was still 500ish a year ago) is one good example of how people with lots of money, and no interest in overclocking keep the market price of one cpu high while a much faster one sells for less (c2d).

[savantu]

Considering recent Wall Street pressure on Intel and the upside AMD got ; Intel will push the pedal to the metal in terms of executing its roadmap.Nothing is going to be delayed , in fact I wouldn't be surprised to see product pull-ins.

As hard as it might seem , AMD manages to hold itself pretty well under Intel's attack.They have a grasp on the low-end market with their <$100 DC and their goodwill with server customers is far more than expected.
Intel really needs to literally obliterate AMD in the >$100 segment in order to increase ASPs , revenue and margins.That's the job for Penryn/Nehalem.

At the same time , Silverthorne will attack the low end market ; although cheap it will be even cheaper to produce => good margins on it.

Thirdly , Larrabee should allow Intel target the mainstream GPU market , extra profit there.

Intel needs to execute on all 3 directions , otherwise it's stock will be hovering in the low '20s and that's certainly not what investors want.

[Shintai]

I am merely saying that Intel isn't pushing nearly as hard as they would be if they were challenged.

Between 2000 to 2008, are you kidding? We were running 1GHz P3s and Athlons in 2000. By 2003 we had 3GHz P4s and 2GHz Athlon XPs.
While the clockspeed of Core2 hasn't increased over P4, it has doubled the IPC of Netburst, so effectively a 3GHz Core2 is equal to a 6GHz P4, and I'm not even taking into account the extra core.

Competition drives innovation, and if Intel isn't challenged, there is less incentive to innovate. Just look at nVidia as a classic example, they have had the 8800GTX for nearly 18 months, because of lack of competition from ATI/AMD they have dramatically slowed down their development cycle.

So before we needed 3 years just for twice the performance? And you already nag? LOL....seriously..get a grip.

Even if a 3Ghz matches a 6Ghz P4. They cant be compared in such a way. P4 reached 3.8Ghz in retail. Also you make a fine example. 2Ghz XPs 5 years ago. AMD got what today, 3.2Ghz X2s besides phenom. You are trying to make an imaginary illusion of technology progress is slower today, yet its far faster.

[adamsleath]

You are trying to make an imaginary illusion of technology progress is slower today

it has yet to be proven that progress is faster "today"

anyway time will tell; and it totally depends how you measure 'progress'...superpi time?:rolleyes:
if 45nm will be the thang for the next 2 years in cpu's then MHz (or GHz) limits theoretically will remain stuck, limited by whatever the best stepping they bring out in 45nm.

so either fancy-shmancy architectural changes and or smarter multicore-like programming is needed to maintain the idea/fact of "progress" in cpu 'power'

[Shintai]

We will et nehalem on 45nm to help offset it. And then 32nm to kick it forward again. Also you forget the 45nm CPUs are faster clock for clock than their 65nm cousins.

[adamsleath]

Also you forget the 45nm CPUs are faster clock for clock than their 65nm cousins.archi. changes as i said. nehalem more archi changes. wot the tweaks are besides cache i do not know exact details...but in my ignorance i say "archi. changes" to refer to design changes in cpu ic that results in increased performance clock for clock.
more instructions per clock or somesuch.

[duploxxx]

As hard as it might seem , AMD manages to hold itself pretty well under Intel's attack.They have a grasp on the low-end market with their <$100 DC and their goodwill with server customers is far more than expected.


join the Shintai club, you both have no idea how the server world is working.... goodwill with server customers :rofl: :rofl: :ROTF: :ROTF: :rofl: :rofl:

check price/performance/power use and you know that depending on the application one cpu firm is just better then the other one. The problem is that most former IT heads just see the name, not the facts and this is changing more and more...

it's just a few days ago that Shintai replied that the server share was decreasing a lot for AMD .... well guess what, it increased again, and just with a small launch of quads and 0&#37; OEM suppliers using them... check again when OEM start sales in March....


We will et nehalem on 45nm to help offset it. And then 32nm to kick it forward again. Also you forget the 45nm CPUs are faster clock for clock than their 65nm cousins.

clock for clock only faster when they benefit from increased cache,fsb bump and sse4, so yes Intel did a good job on their shrink and power consumption now it has to be proven that they will do well on Nehalem........ offcourse you already now they will do well.

[AMDDeathstar]

Intel could easily release a 3.8GHz quad TODAY and remain under 130W TDP.

I'm sure they could make such a chip but how much would it cost. Local shop here just listed QX9650 for 1,129 dollars. Even if wasn't EE I don't see it being cheaper. You'll be paying a hugh premium for chip has to be stable at stock speed of 3.8GHz for 3 years(lenght of warranty).

[Donnie27]

I'm sure they could make such a chip but how much would it cost. Local shop here just listed QX9650 for 1,129 dollars. Even if wasn't EE I don't see it being cheaper. You'll be paying a hugh premium for chip has to be stable at stock speed of 3.8GHz for 3 years(lenght of warranty).

Same price as the chips that overclock to that and beyond. It's not like the only want to get 3.8GHz is to buy one;) AMD has to worry about those $266 2.66GHz 45nm models in or about April. Not the Ep0s models folks noobs buy! Oh and Intel will sell plenty of them for $1,200 to $1,300. Just as folks paid that much for FX-57 single cores LOL!

[Epsilon84]

I'm sure they could make such a chip but how much would it cost. Local shop here just listed QX9650 for 1,129 dollars. Even if wasn't EE I don't see it being cheaper. You'll be paying a hugh premium for chip has to be stable at stock speed of 3.8GHz for 3 years(lenght of warranty).

Well, AMD pushes an X2 6400+ to the absolute limits of K8 and it costs $170... food for thought on that 'huge premium' theory. It all comes down to what the competition is delivering.

So I guess it will cost as much as the market can bare. If AMD actually had a competitive CPU out you can be sure a 3GHz Penryn wouldn't cost over $1000 right now...

[Poodle]

Intel wouldn't dealy 45nm if they didn't have to..

[Shintai]

Well, AMD pushes an X2 6400+ to the absolute limits of K8 and it costs $170... food for thought on that 'huge premium' theory. It all comes down to what the competition is delivering.

So I guess it will cost as much as the market can bare. If AMD actually had a competitive CPU out you can be sure a 3GHz Penryn wouldn't cost over $1000 right now...

Its abit of an unequal comaprision. AMD selling 3.2Ghz X2s for 170$ left them where? And with what money for factory upgrades and R&D?

Its abit like peeing your pants as a consumer. Great now for good prices. But not so great in the future.

[JohnZS]

My QX9650 is still at RMA :confused:
Anyway on topic nobody here has mentioned that the retail Q9*50 CPU's are actually a different stepping to the QX9650 and Xeon 45nm (which would explain why those 2 CPU's are not effected).


SLAWQ C1
12 MB
(2x6 MB)
10677h Q9550
2.83 GHz /
1333 MHz
775-land LGA

And the source of this information is directly from Intel Intel QX9650 and Q9000 Errata Document (http://download.intel.com/design/processor/specupdt/31872704.pdf)

Review samples are ES chips based on the C0 SLAN3 stepping (same as the QX9650 and Xenon chips), however the retail 9000 seris are based on the SLAWQ C1 stepping.

John