"Stress technology" can you say K9 4Ghz in 2007.

Serge84

Gawd
Joined
Nov 5, 2005
Messages
693
http://sharikou.blogspot.com/2006/03/intel-to-suffer-severely-from-osborne.html

http://www.amd.com/us-en/Corporate/VirtualPressRoom/0,,51_104_543_13743~103048,00.html

Sorry but this is quite intresting AMD fans will love AMD again and forget about conroe. Yeah I noticed I spelled it wrong. >_> Sorry about that.

When AMD goes to K9 All hell will brake loose. The new arc will perform at speeds of 4Ghz around 2007. I knew something like this was coming but AMD is very tight liped about their K9's. They where prepared for this for 2 years. Don't be surprised. And 4ghz isn't impossible for AMD even now, they just been waiting for the right time and wow its coming. Very scarry shit. Do you know what this makes conroe look like now... damn. I'll try to find more tech on it but if AMD fans new about it then why is nobody saying anything about it. No wonder AMD isn't even surprised about conroe... nore flinched one bit.

Also quad core opterons will be in rumored 4ghz vers very soon. As said in a artical and news itself. This is more of what the socket is capable of as I stated. Since its of next gen parts. K8L wasn't sipose to be impressave K9's are. Everybody gets over excited for nothing ya know. AM2 wasn't all about K8L but was about K9's. Thats getting to the beginning of the real potentale of the socket when it matures some such as here. This is faster then I ever thought. But it is still with in a reasonable time frame to the law of speed in the comp world. This roughly makes amd's 40% faster then K8's this hurts Conroe badly.

There are few articals mensioning this large speed leap AMD announced. This is one of the techs they been so tight liped about. Now we know. Woot. Go AMD. Socket F and AM2 get 4ghz or near 4ghz after the new arc is applyed at Fab36 with the use of 65nm's. K8L's will be one of the shortest lived rev or arc mods ever in history. K8L's appear to only start off AM2 into the market place and when 2007 comes and K8L's are replaced with K9's. This will jump AMD back into the lead very soon after conroe. I hope AMD has made K9's kick ass. Sounds like I'm going to call my CPU a dog now. XD This will be quite impressave at that.

I'd like to know more about this tech tho. Just sounds so bad ass. When AMD doesn't hold back they sure don't hold back damn. Now thats impressave. 4ghz AMD stock speed. o_O SO soon... bloody hell. Let the dogs out!!!

It will be deja vu all over again. http://sharikou.blogspot.com/2006/01/intels-next-generation-merom-may-have.html
 
I hate to be the one who says anything..... but......

I think you might be a liitle too enthusiastic. AMD pretty much stopped marketing on the Kx nomenclature.... K8L=Rev. G

Your talking about Rev H parts that we wont see untill late 07 at the earliest. Personally I think we need to worry about Rev F first, and lets not get too far ahead of ourselves. I believe that AMD will lose the desktop crown with Rev F... But not by a huge margin... They will still be able to compete....

Rev G will prolly take the lead back, but that is still to be seen. It is still to soon to speculate.

Also are you talkin about dual stress liner? If so, then it is important to note that AMD has neen using it in limited quantity since the FX55. But yeilds have not been too great so they have not fabbed much with it. They say that it will be standard on all 65nm parts. But that is speculation.
 
JackPack said:
Link to Sharikou.... All credibility lost.

yep. That's where I stopped reading. There are some decent roadmaps for AMD's quad core parts on HKepc
I think I'll stear well clear of the impending train wreck.

I will however grab: a lawnchair, a beer and an asbestos suit and enjoy the carnage.
 
FreiDOg said:
yep. That's where I stopped reading. There are some decent roadmaps for AMD's quad core parts on HKepc
I think I'll stear well clear of the impending train wreck.

I will however grab: a lawnchair, a beer and an asbestos suit and enjoy the carnage.

and i'll bring a 6pack and watch next to you.
 
Firt of all k8l is in early 2007. Second 4 gighurtz huh, how do you get 4 ghz from the amd article... Did you look at the sige 65 nm chart... Germanium (e-SiGe) with Dual Stress Liner (DSL) and Stress Memorization technology (SMT) on Silicon-On-Insulator (SOI) wafers, resulting in a 40 percent increase in transistor performance compared to similar chips produced without stress technology, while controlling power consumption and heat dissipation Do you realize that the more things they do to regular silicon the more expensive the silicon might be?
Also think of this rationally, maybe they are doing this because they might not be able to get DECENT yeilds without it...
3rd until ht 3 technology is on the processor (in any form that we see), I wouldn't hype it. 4th Please if your going to post sharku read and try to understand a few more point of views.
 
duby229 said:
I hate to be the one who says anything..... but......

I think you might be a liitle too enthusiastic. AMD pretty much stopped marketing on the Kx nomenclature.... K8L=Rev. G

Your talking about Rev H parts that we wont see untill late 07 at the earliest. Personally I think we need to worry about Rev F first, and lets not get too far ahead of ourselves. I believe that AMD will lose the desktop crown with Rev F... But not by a huge margin... They will still be able to compete....

Rev G will prolly take the lead back, but that is still to be seen. It is still to soon to speculate.

Also are you talkin about dual stress liner? If so, then it is important to note that AMD has neen using it in limited quantity since the FX55. But yeilds have not been too great so they have not fabbed much with it. They say that it will be standard on all 65nm parts. But that is speculation.

Ahh yes thats true. But still quite impressave at that. This is just like the Ghz wars all over again. Coming pritty close. I been doing some research. AMD has done upto K11 desines. Wow. Thats working quite ahead of things Aready planning. But yes it is way too early to say for sure. However its true we will really see the Ghz really rise very high in the next 2 years from now. Should be quite intresting. If they do include this dual stress liner in 65nm parts it will give AMD a great advantage in speed and performance. AMD and IBM are in it together.

And if this starts at mass production with rev H. That can nearly double Ghz speeds. This will if they did it as soon as 65nm tech as AMD and IBM clam and it is hard desined in the arc this will give AMD a great advantage over intel. And knowing AMD's style as well as how they reacted twords conroe. "Says they arn't worried." Im pritty sure they will do it as soon as possibe and AMD isn't one to go canceling things or failing for that matter. They always do what they say and have been that way for years. And I hope they are feeling that compeditive. lol We all want to see some good old wars with AMD and intel just like the old days now don't we.

Yet we don't even know any specs at all on rev G yet. But these are apperently very large leaps. This tech has been around for awile after all tho. They could technicly release it when ever they feel intel is abit ahead. AMD finally starts to slowly show off its true color. But it is sipose to be release as soon as 65nm comes out because its part of the desine. So this will be quite the impressave show. They do have alot of tech they been holding back on for the past 2 years and all. So I wouldn't be surprised if they started skipping rev's. They are just too tight liped for my liking.
 
thecoldanddarkone said:
Firt of all k8l is in early 2007. Second 4 gighurtz huh, how do you get 4 ghz from the amd article... Did you look at the sige 65 nm chart... Germanium (e-SiGe) with Dual Stress Liner (DSL) and Stress Memorization technology (SMT) on Silicon-On-Insulator (SOI) wafers, resulting in a 40 percent increase in transistor performance compared to similar chips produced without stress technology, while controlling power consumption and heat dissipation Do you realize that the more things they do to regular silicon the more expensive the silicon might be?
Also think of this rationally, maybe they are doing this because they might not be able to get DECENT yeilds without it...
3rd until ht 3 technology is on the processor (in any form that we see), I wouldn't hype it. 4th Please if your going to post sharku read and try to understand a few more point of views.

Well Socket AM2 comes out in this month btw K8L's are May 23. As well as Socket S and F I beleave. There is nothing lower then Rev F. lol Its only made to last till 2007 or earlier. AMD said they don't need fab36 to go 65nm's and plan to launch the new arc as soon as possible like at the end of 2006 if possible. They been talking about this, its quite new tho. A matter of days new. They are prob mass producing them as we speak so they can get a early start on getting to 65nms, (Mass produced that is) I mean they aready been having 65nm proto types. They just need to get use to makeing it a a large volume. Not too difficult. They say it will be their fastest transision yet from prosesses. But its a small long start. Maybe even around just in time for conroe knowing AMD. They are under estimated too much. They can do things faster then intel likes. heh And they hold too many damn secrets.

All acording to mainly a law of speed long since said has been right for the past 30 something years as I said before. This law has never been wrong once. We can gane speed in 2 ways be cycle per clock or by raw ghz. K9's K10's and K11's are all new arks and are diff then K8 desines. Amd for one has been too tight liped unfortionatly so nothing can be said for sure. But every 2 years clock speed doubles eather way. By 2008 we should be at 8ghz. This isn't the limit of silicon but the limit of arc. K8's can't generate the speed but a new arc could. Conroe could do the same actouly it does per cycle that is. So its 10% faster then a K8. Oh and I seen the speeds talked about in afew articals to be 4ghz or close enough to it. 3.8ghz and so on. Sure on K8L its impossible but like I said if a new arc is released, (Witch it is) said by AMD themselfs they release their new arc as soon as they move to 65nm's. K9's or K10's will be the shit. And I'm sure intel has something like this in mind for their cpus as well with more insane cycles per clock.

K8's are 3 year old arcs. They have done something for the past 2 years of that in secret We all know intel and AMD make next gen CPU's that kick ass when a new arc is interduced. Rev F isn't sipose to last long. Its only a upgraded K8 as K8L made to get AM2 started for the most part. The real power here is the 65nm arc new arc at that AMD and IBM clam are really something but again its too tight liped. 4ghz isn't impossible now. Infact in reality if we mesured a AMD speed wise to a P4 it would be running at 4.2ghz or so. At a speed of 2.8ghz. Well thats what sandra tells me my speed is compared to a P4 at 3.8ghz. My rateing is 12700 in sandra. Wile a P4 running 3.8ghz runs only 11000. Thats quite afew hundred mhz in speed compared to a P4 because K8's have a much faster cycle per clock a 2.66ghz conroe will be like a 4.5ghz equivlelent to a P4 CPU.

But getting raw speed through a new arc just like conroe. K9 or whatever they deside to use. Some say they skipped to K10. Well whatever it is. It costs much cheaper to make CPU's now adays as the tech gets more advanced and utalize techniques or improvements they invent who knows. You know they have always been compeditave. So Intel isn't the only one who has their bad ass processor. Yet AMD has always been that way. They are like the scilent killers. They strike when only they feel like it. Always has been that way since the K7 days. Remember Rev A and Rev B. What a difference. Sames going to be from Rev F to Rev G and Rev H. According to how AMD's style is. Look how bad ass we think our current revs are such as E's. Where all the bomb back before conroe. AMD still leads for now. If you don't include conroe.

****FreiDOg 2[H]4U
FreiDOg is online now Report Bad Post
Quote:
Originally Posted by JackPack
Link to Sharikou.... All credibility lost.

yep. That's where I stopped reading. There are some decent roadmaps for AMD's quad core parts on HKepc
I think I'll stear well clear of the impending train wreck.

I will however grab: a lawnchair, a beer and an asbestos suit and enjoy the carnage.****

Well isn't there a better road map say ones for Desktops only? And this could change at anytime really. Just like AM2 comes out 2 weeks ahead of time.
 
Sigh. I stopped reading and gouged out my eyes when I came upon something "intresting".
 
banGerprawN said:
Sigh. I stopped reading and gouged out my eyes when I came upon something "intresting".

Just hope AMD doesn't release 65nm tech as they want to ahead of time. For your sake the 1st day of 2007. =P AMD isn't playing around anymore since they don't need fab36 to make 65nm tech. They can do it with current tech at fab30 and say it could really cut the time in half. Things have changed when the news came in April 26th I beleave it was. AMD announched it. lol Scary how fast they can do things how ahead of the norm they originaly thought.
 
Serge84 said:
Well isn't there a better road map say ones for Desktops only? And this could change at anytime really. Just like AM2 comes out 2 weeks ahead of time.

The HKepc roadmap is the first detailed look we've had at AMD's quad core roadmap, I don't think there are any credible reports that differ substantively from that.

Roadmaps are subject to change as the market, or more likely the engineers, require. AM2 was supposed to ship in April, but trouble with the memory controller bumped it to June, and now for whatever reason AMD has pushed it up two weeks. The Turion X2s have slid from their May release because nobody wants the current Turion chips. AMD doesn't want a big Turion X2 launch if they haven't found anyone to buy them.

That said, I doubt very much Greyhound will move up very much. As 'cool' as they may be to enthusiasts, quad core parts serve little purpose on the desktop. Considering how slowly software generally responds to new technologies, I would not expect there to be significant demand for consumer quad core parts in the next year to year and a half.

Out of curiosity, is English not your native language? I don't like to pick on people online for bad spelling or grammar (mostly because of my bad spelling and grammar), but it's making your posts difficult at best to follow. You might want to run your posts through a spell checker before actually submitting them.
 
FreiDOg said:
Out of curiosity, is English not your native language? I don't like to pick on people online for bad spelling or grammar (mostly because of my bad spelling and grammar), but it's making your posts difficult at best to follow. You might want to run your posts through a spell checker before actually submitting them.
I was just about to say the same thing. :D
 
4GHz from a company whose products sell on IPC? I wouldn't bet on it. Intel are struggling their butts off to get 4GHz out of netburst within a sane thermal and electrical envelope, which was DESIGNED and INTENDED to scale to those sorts of speeds.

Hell no. I'm sure the technical hurdles will be overcome to have a >4GHz proc with the same sort of IPC of the K8, with a low power consumption, too, but NOT NEXT YEAR.

This lot sounds like market FUDling by AMD, if you ask me. And no, I'm not an intel phanbooi.
 
xX_Jack_Carver_Xx said:
I thought the Toledo's, Sandiego's, etc were already stressed-silicon on silicon-on-insulator.

Not all of them are. AMD has some produced using Stress liner... But most are not. I havn't figured out a good way to tell which is which though.

My guess though, is those chips that make it to 3 ghz are, and those that dont aren't.
 
SS on SOI is interesting but i thought dual gates was AMD IBM next big performance jump. Interesting thread . But a couple of links would be nice.( differant than the ones you posted)

I tried saying 4 GHz on K9 but I had a lot of trouble doing so without dropping to all fores and started woofing. lol.
 
$BangforThe$ said:
SS on SOI is interesting but i thought dual gates was AMD IBM next big performance jump. Interesting thread . But a couple of links would be nice.( differant than the ones you posted)

I tried saying 4 GHz on K9 but I had a lot of trouble doing so without dropping to all fores and started woofing. lol.

Dual gates are still a little ways off. That is a toy for the carbon nano tube generation. They need to figure out a way to mass produce a good carbon gate pad first. Once that is done however.... 3d processes should be right around the corner.....
 
perplex said:
haha you are so funny. Seriously, how old are you?

Old enough to reason out that K9 @ 4 ghz is very funny. Also I thought AMD was going to Jump over the Doggy core.
 
duby229 said:
Dual gates are still a little ways off. That is a toy for the carbon nano tube generation. They need to figure out a way to mass produce a good carbon gate pad first. Once that is done however.... 3d processes should be right around the corner.....

Intel will be going to 3d gates on the 32nm die . Why does AMD IBM have to wait on nano gates? High K with metal gates should do the Job.
 
$BangforThe$ said:
Intel will be going to 3d gates on the 32nm die . Why does AMD IBM have to wait on nano gates? High K with metal gates should do the Job.

they don't, and they won't.
IBM, AMD, and TSMC are targeting either 45 or 32nm for widespread use of FinFETs (a.k.a. Double Gate transistors ).
IBM has been experimenting with them in SRAMs for years now, implementing them is probably more a matter of financial practicality than technological impossibility. FinFETs don't require substantial (in comparison to Tri-gates) changes to the manufacturing process so they may end up being easier to implement. Though Tri-gates look like they have some functional advantages over FinFETs.
 
AMD IS NOT sitting on its hands in the face of the Intel Merom attack.

AMD is not talking about releasing QC parts until G, but sources tell us if sales start flagging, or if Intel starts kicking them around, it could be pulled in to F this year. It is a marketing thing, just like DDR2 on E was, and of late, AMD seems to be very good at timing.

http://www.theinquirer.net/?article=30042

So AMD has Rev F quad core parts. Ahh intresting according to the sources anyways.

This is also intresting. But really now. Why is 4ghz so hard to beleave. We aready passed it, clock per clock in cycles really. Pure clock speed is much faster then it was a year ago, and it could be lower then something was a year ago clock per clock but cycles make it even faster so infact a FX2.8Ghz is really running at +4.2Ghz equivlent. We will see much faster clocks from Rev G. Rev F shouldn't even count. Its only like a souped up Rev E and is still K8 based. This K8L is really sad. K10 should be much nicer even if a 20% only in speed is made. Thats atleast 3.5ghz stock with a increase in speed like that. But its too early to say.

We are sipose to be doubling clock speed every 2 years. They still are understand. This will happen. As they reach limits as they did to get past 1ghz they break them. Net Burst at 4ghz couldn't be made because it lost funding and was a failure. They tryed to push the tech to the limit. It could do it in a OC. Enough time they could have really done it at 4ghz but they just said lets move on to something else. Thats why P4 4ghz's where canceled. Just like others after. AMD is completly diff, they don't cancel things its not like them and they don't fail, maybe one day, but they don't make the same mistakes like intel. Trust me your going to see 4ghz or even 5ghz in 2007 sometime from AMD. Yet they are too tight liped for their own good. Ppl think they been doing nothing when they really have impressave stuff they just don't tell anybody about. AMD is so kick ass for a reason. And with IBM this will be quite intresting. We are just going to have to see what happens in the end. But we must never under estimate AMD.

Unless the speed of law that has been going on for 30 years actouly falls. That will be a 1st. How many times do I have to use that. It simply has never been wrong once "." This thrend was only made for speculation so don't take anything so seriously. Its speculation on what AMD is planning for one with the K9/10/11 arcs.
 
Serge84 said:
We are sipose to be doubling clock speed every 2 years.<snip>

Unless the speed of law that has been going on for 30 years actouly falls.
it's nothing about the speed, but the number of transistors on a die.. ;)
 
Serge84 said:
We are sipose to be doubling clock speed every 2 years. They still are understand. This will happen. As they reach limits as they did to get past 1ghz they break them. Net Burst at 4ghz couldn't be made because it lost funding and was a failure. They tryed to push the tech to the limit. It could do it in a OC. Enough time they could have really done it at 4ghz but they just said lets move on to something else. Thats why P4 4ghz's where canceled. Just like others after. AMD is completly diff, they don't cancel things its not like them and they don't fail, maybe one day, but they don't make the same mistakes like intel. Trust me your going to see 4ghz or even 5ghz in 2007 sometime from AMD. Yet they are too tight liped for their own good. Ppl think they been doing nothing when they really have impressave stuff they just don't tell anybody about. AMD is so kick ass for a reason. And with IBM this will be quite intresting. We are just going to have to see what happens in the end. But we must never under estimate AMD.

Unless the speed of law that has been going on for 30 years actouly falls. That will be a 1st. How many times do I have to use that. It simply has never been wrong once "." This thrend was only made for speculation so don't take anything seriously.

if you're refering to Moore's Law, it's not about doubling clock speed. http://en.wikipedia.org/wiki/Moore's_law - for the love of God please read this

edit: and to say AMD doesn't fail is laughable. I have yet to see a perfect company that doesn't fail at something.
 
duby229 said:
Not all of them are. AMD has some produced using Stress liner... But most are not. I havn't figured out a good way to tell which is which though.

My guess though, is those chips that make it to 3 ghz are, and those that dont aren't.

They need 65nm's to do it in massave builds. If they take advantage of this using their current way they make the core on a SOI "Strained silicon" and a diff arc that will do the trick. This is what AMD has been braging about a bit. 65nm's will allow more space and transisters to work with. On 300 warfers this will be quite intresting if SL is really is standard. This may make the prosess much cheaper if they mass produce it. Its so expencive because they don't mass produce it for one. And because its harder to include on 90nm's then it is with 65nm's its much easyer. They say because of the new arc they can do it with 65nms too. It is not hard to double transisters eather. Its just more expencive. If they got the money then this will really be something.

Just all depends how good this new K10 or whatever arc will be. The desine has everything to do with how good they will be. If they use Zram L3. Made L3 much smaller in cells. Made a larger core and compacted the memory controler and other compoments. We will see a vast difference in transister count and performance. I'd like to see them have as much transisters as a ATI GPU or more. lol That would cost a fortion and make more heat. But AMD knows how to make CPU's better then anybody. So I wouldn't be surprised if they do or don't. Its not a matter if tech. Cus they know how to do it. They just need the money to do it. If thay find new techniques in making this stuff cheaper then we won't see them hold back much longer. I hate they wait all calmly to do crap. And intel rushes things out so fast. They aren't nearly as tight liped as AMD is. That can get on the nurves of a modder.

AMD=Many years with no failures and dominated for 3 years with K8's. Intel=3 years of failures with P4 P5 and netburst as well as many others and alot of cancelations. AMD has yet to recently fail or cancel anything. (I never new they did). They have a nearly perfect record. It would be a joke to say Intel has ever done anything right for the past few years. Sure somethings but others not so much. Intel is a mistake. AMD's are more reliable as why most goto them for stuff. Intel is so big because they are more wide spread and have 10 times the employees AMD has. But AMD could take a good bit of intels shares by the end of this year.

And the law is about performance and advancement. Same thing I ment since speed equals performance. Every 2 years tech gets twice it was before that. But they mainly mean speed. The law you stated is stateing the wrong area.
 
What I'm thinking. I find your lack of faith..... disturbing.


200210.vader1.240.jpg


Don't be a !!!!!!.
 
PureBooYah said:
if you're refering to Moore's Law, it's not about doubling clock speed. http://en.wikipedia.org/wiki/Moore's_law - for the love of God please read this

edit: and to say AMD doesn't fail is laughable. I have yet to see a perfect company that doesn't fail at something.


http://lowendmac.com/musings/buynew.html

This is the real Moore's Law. You posted a incorrect matter.

Moore's Law

Moore's Law states that the number of transistors in a given area will double every 18 to 24 months. What that means in the real world is that memory chips double in capacity every 18 months or so and CPUs pretty much double their performance on the same schedule. Likewise, although not related to Moore's Law, we see the same kind of progress with hard drives.

Imagine that you bought a new Suburban in 1999 that could just hit 55 MPH on the open road, guzzled gas at 10 miles per gallon, and cost you $30,000. If cars followed Moore's Law, the 2001 Suburban would top 120 MPH, get around 25 MPG, and cost you $20,000. Wouldn't you feel hopelessly outclassed if all the 2000 cars were zipping past you at 85-90 MPH and the 2001s screaming past at over twice your speed?
 
texuspete00 said:
What I'm thinking. I find your lack of faith..... disturbing.


200210.vader1.240.jpg


Don't be a !!!!!!.

XDD I joined the darkside along time ago. :cool: "AMD" Faith is only in the eye of the beholder. And I beleave AMD will do everything they say they are. Because they always have. My faith is strong in the darkside. lol
 
Serge84 said:
http://lowendmac.com/musings/buynew.html

This is the real Moore's Law. You posted a incorrect matter.

Moore's Law

Moore's Law states that the number of transistors in a given area will double every 18 to 24 months. What that means in the real world is that memory chips double in capacity every 18 months or so and CPUs pretty much double their performance on the same schedule. Likewise, although not related to Moore's Law, we see the same kind of progress with hard drives.
yes, and you explicitly stated that speeds double, not transistor density.. ;)
 
Heh, the darkside is different for everyone I suppose. I thought Intel was the darkside? :p The empire versus the rebellion. Makes sense. Intel is the darkside. I specifically picked the quote and pic where he had the hand out. It is indeed the dark side conroe choking the crap out of you for disbelief. :D

Just seems a lot of wishful thinking been going around on your part. All so AMD can avoid a phase all chip makers go through. That being the performance second fiddle. As a 5 year + AMD user, part of my heart sank when I first saw the kick in the teeth coming. Then I realize it's really not that important. We are people, AMD is a stock, with some fabs in Germany.

At least half the reason you start these threads is the performance in Conroe. It's ok. Let it go.
 
This thread is worthless without .. wait, it's just plain worthless.

I am amused by the f4nbois rambling on about sujects they know nothing about. It's amusing.

Have fun when the reality kicks in (and I don't mean that AMD can't currently compete).

*takes off asbestos suit and goes swimming*

Someone throw me a beer.
 
mwarps said:
Someone throw me a beer.
*toss* :p

also, the rumor posted before about some of the revE's being SSOI and some of them being just plain SOI.. if the former increases yields and transistor performance, why make anything without it? ;)
 
do you amd !!!!!!s ever give up?

AMD could be bankrupt and their fabs burn down... !!!!!!s would just say "oh its their new strategy to take on intel!!" "now they can build from the ground up and retake the crown!"

lol :)
 
Back
Top