Anand re-examines Conroe benchmarks

ahh.. well.. that would make perfect sense then..

lets see.. FAB 36 is a 300mm waffer facility.. (IIRC), so... each peice of silicon produces 111111111111 chips at 90nm or 213017751479 at 65nm.. thats almost double the amount of chips.. double the money made...


... did i do that right?
 
general habbit is that people buy whats the best at that time so obviously the meaning of "best" only lasts several months before something new comes out. However if you plan on buying a new system you might as well wait for conroe (if you are an intel user) or AM2 (if you are AMD) so that you know that your rig will not be obsolete in such a short time.
 
--

All right. Since you obviously can't listen to reason and common sense, I'll post a link to a simiconductor fabrication site.

http://www.fabtech.org/index.php?option=content&task=view&id=1186&Itemid=2
The devices being fabricated at Fab36 use AMD's 65nm process but on 90nm design rules to assist in a fast ramp with high yields.
Click to expand...

--

empoy said:
so can you explain how are you going to fabricate 90nm Si dies with 65nm equipment? Are you kidding me? are you working for a semicon company?
Click to expand...
Listen, I said the same fucking thing two fucking posts in a row. By the industry standard, a process is described by the smallest feature it can produce. 65nm doesn't mean everything is 65nm, it means that the smallest feature is 65nm. Any size over 65nm should be possible on 65nm equipment.

I'll say it one more fucking time: Eat some of your own advice and google a little, because you don't know what the fuck you're talking about.

--

Narisatu said:
Im just curious.. but why CANT you fabricate 90nm dies with 65nm equipment? i wouldnt think it would be that hard to use smaller equipment to make something bigger... though.. that just answered my question i think.. it would be like using a dremel to cut down a tree....
Click to expand...
Making CPUs does not involve physical tools like that. It's not like you need a tiny drill to make tiny holes, and only one size of hole can be made with one drill-bit of a given size. Making CPUs involves shining short-wavelength light through a number of "masks" which block out portions and cause complicated interference patters with others. While the whole story is a fuck of a lot more complicated than what I describe here, all one really needs to do is switch to a different set of masks to make processors at different nm sizes. When one says that they have a 65nm process, that means that one has the masks and other gear that is accurate down to 65nm. That's accurate to 65nm. Similar to the concept that a number that is accurate to 100 decimal places is also accurate to 10 decimal places, the 65nm process can make anything less detailed, or larger, than 65nm.

Again, this isn't a hard concept...


--

Bao01 said:
Whatever he's about to say now, I hope he understands the right meaning whenever he finds out about AMD "tooling up" for 65nm.
Click to expand...
Again, more posts from people who want to pretend like they know something. AMD has 65nm gear right now, and is making working silicon on it. AMD is selling 90nm CPUs off this line right now, and will transition the equptment they already have to 65nm as soon as they decide they can and they need to.

--

Let me say this for the benifit of everyone:

If you honestly know what you are talking about, post away. If you don't know what the hell you are talking about, don't pretend to. You know what that's called? Honesty! I swear, it feels like I'm in a highschool crowd full of little teens trying to bullshit all over the place to impress people: I find it horribly pathetic and depressing.

One of these days I'm going to leave all the "[H]ard" kids to bullshit among themselves and just give up on this forum.
 
Conroe looks to be an absolute monster. The beating it gives AMD is reminiscent of A64 vs p4. GG intel.
 
One of these days I'm going to leave all the "[H]ard" kids to bullshit among themselves and just give up on this forum.
Click to expand...

LOL! Touchy aren't you? Either humor is not your best friend or you've just pissed your pants because you've uncovered something new. I can understand if your temperment makes forum life difficult for you.
 
Bao01 said:
LOL! Touchy aren't you? Either humor is not your best friend or you've just pissed your pants because you've uncovered something new. I can understand if your temperment makes forum life difficult for you.
Click to expand...

Well, he could be touchy due to the fact that everyone has attempted to rebuke what he says with no facts or anything, they then attempt to change the subject after he provides fact.
 
And why is it you Intel people can't compare current product to current product? So many people have complained about this, one would think you would get it by now. I don't care about AM2 or Conroe right now:

Intel promised 10GHz P4s.
Intel promised the Itanium would take over the CPU industry with amazing performance.
Click to expand...

The difference would be that all that was just on paper. There is shit load of Conroes out and all of them are kicking ass. There was no benches of the Itaniums kickin ass either so I never really believed that. Again ....believe nothing that you hear and everything that you see. And honestly we have seen enough.


Sure Conroe is not current, but its in the bag and I honestly think it would be out sooner if Intel didnt just release the Presler.


Right now, all I care about is the fact that AMD is kicking Intel's ass. Six months from now I may be saying the opposite.
Click to expand...


I dont know about kicking ass. My Presler is actually faster in certain games @ lower than 16x12 resolutions , but most of the time its not that much slower. Q4 has the biggest difference of around 20%.

I have not benched much, but from all the reviews the 955XE spanked AMD in certain benchmarks so Id say AMD has just a slight lead. Slight spanking.....butt cheeks are just a bit rosy.
Im not talking only games as there is more to computing than just games.
Im not overclocked yet as I need to finish some things with my system first, but I think Ill be pretty happy once the clock is raised

Sure new hardware comes out every 6 months , but how often do you see benches of something that will be out in 6 months. Not to mention that those were not just internal benches. They actually had a bit of control over the way they benched.

I have never seen benches from ATI,nVidia,AMD or Intel six months before release. All I heard was promises just like you. Now wouldnt you say that this situation is a bit different????

Im glad that you acknowledge that in six months you might be saying something else which is more than most of the people that base their argument on the "current" theme.
 
Bao01 said:
LOL! Touchy aren't you? Either humor is not your best friend or you've just pissed your pants because you've uncovered something new. I can understand if your temperment makes forum life difficult for you.
Click to expand...

That is true. I am a bit touchy... and grouchy...

But come on, I posted the same freaking thing three times, and I bet I will still get responses from people saying: "WTF? 90 is not the same number as 65!! It won't work!"
 
visaris said:
That is true. I am a bit touchy... and grouchy...

But come on, I posted the same freaking thing three times, and I bet I will still get responses from people saying: "WTF? 90 is not the same number as 65!! It won't work!"
Click to expand...

There you go again. You were getting frustrated by one guy's snippy comments, a joke, a smiley, some metaphors and suddenly the whole world is after you. Come on, man :)
 
Bao01 said:
There you go again. You were getting frustrated by one guy's snippy comments, a joke, a smiley, some metaphors and suddenly the whole world is after you. Come on, man :)
Click to expand...

Yes, and it's easy to get frustrated when people are running their mouths about stuff that they have no idea about. Happens all the time. You're not any better by playing mr. forum mediator. That's why we have mods here. :) And since I put a smiley after this, you're not supposed to come back with any retorts because it's all in fun :)
 
D4hPr0 said:
I dont know about kicking ass. My Presler is actually faster in certain games @ lower than 16x12 resolutions , but most of the time its not that much slower. Q4 has the biggest difference of around 20%.

.
Click to expand...


Are you joking? You're at stock and saying that AMD is barely faster than your 955, IN GAMES!? lol. :eek:
 
sabrewolf732 said:
Are you joking? You're at stock and saying that AMD is barely faster than your 955, IN GAMES!? lol. :eek:
Click to expand...


I said in certain games. I know AMD owns most games....Im an Intel guy, but Im not stupid.

This is the best I could come up with right now, but I wouldnt say that the 955EE gets its ass kicked......840EE most definitely though as it was a joke of a processor.
 
While visaris' story is right from a technical point-of-view (65 nm equipment can build chips at pretty much any scale, down to 65 nm), technically AMD does not have a complete 65 nm process yet.
That is, they have 65 nm equipment, but either they don't have all parts of the production-line ready for 65 nm yet, or they are having problems with scaling their products down to 65 nm... because fact is, AMD is not producing 65 nm processors yet.
And it's not a good thing to invest in a 65 nm production-line, and continue to produce 90 nm chips, because you won't get a return on your investment.
I think AMD is going to have a very rough time when Conroe comes out. In fact, they may already be having a rough time, but it's not that apparent to the public when their chips are still performing well.
For Intel it was the other way around, lately. Intel had everything together with the 65 nm process, and their sales figures and profits were fine aswell... But the focus was on the 'bad' performance of the P4.
 
visaris said:
--

All right. Since you obviously can't listen to reason and common sense, I'll post a link to a simiconductor fabrication site.
....

Listen, I said the same fucking thing two fucking posts in a row. By the industry standard, a process is described by the smallest feature it can produce. 65nm doesn't mean everything is 65nm, it means that the smallest feature is 65nm. Any size over 65nm should be possible on 65nm equipment.

I'll say it one more fucking time: Eat some of your own advice and google a little, because you don't know what the fuck you're talking about.

Making CPUs does not involve physical tools like that. It's not like you need a tiny drill to make tiny holes, and only one size of hole can be made with one drill-bit of a given size. Making CPUs involves shining short-wavelength light through a number of "masks" which block out portions and cause complicated interference patters with others. While the whole story is a fuck of a lot more complicated than what I describe here, all one really needs to do is switch to a different set of masks to make processors at different nm sizes. When one says that they have a 65nm process, that means that one has the masks and other gear that is accurate down to 65nm. That's accurate to 65nm. Similar to the concept that a number that is accurate to 100 decimal places is also accurate to 10 decimal places, the 65nm process can make anything less detailed, or larger, than 65nm.

Again, more posts from people who want to pretend like they know something. AMD has 65nm gear right now, and is making working silicon on it. AMD is selling 90nm CPUs off this line right now, and will transition the equptment they already have to 65nm as soon as they decide they can and they need to.

One of these days I'm going to leave all the "[H]ard" kids to bullshit among themselves and just give up on this forum.
Click to expand...

You Sir , are one big idiot.

Not only that you don't have a clue on chipmaking , but worse you jump on others who guess what , work at Intel , and they have at least an ideea of what chipmaking implies.

The other problem is , you are so gullible that it hurts.

AMD made FAB36 from a production POV purely 90nm.
Only in January this year they applied for a 500mil $ loan to buy 65nm equipment.At best that equipment is installed now.

However, recent reports indicate that AMD is attempting to raise capital in excess of $500 million USD to expand Fab 36 production equipment to support a move to 65nm -- the next big step for fabrication.
Click to expand...

http://www.dailytech.com/article.aspx?newsid=418

You can apologize to empoy if you are as "honest" as you say...but IMO , far from it.
 
Spare your breath, savantu. There's never enough ammo to fight trolls like visaris, unfortunately... :(
 
visaris said:
--

All right. Since you obviously can't listen to reason and common sense, I'll post a link to a simiconductor fabrication site.

http://www.fabtech.org/index.php?option=content&task=view&id=1186&Itemid=2


--


Listen, I said the same fucking thing two fucking posts in a row. By the industry standard, a process is described by the smallest feature it can produce. 65nm doesn't mean everything is 65nm, it means that the smallest feature is 65nm. Any size over 65nm should be possible on 65nm equipment.

I'll say it one more fucking time: Eat some of your own advice and google a little, because you don't know what the fuck you're talking about.

--


Making CPUs does not involve physical tools like that. It's not like you need a tiny drill to make tiny holes, and only one size of hole can be made with one drill-bit of a given size. Making CPUs involves shining short-wavelength light through a number of "masks" which block out portions and cause complicated interference patters with others. While the whole story is a fuck of a lot more complicated than what I describe here, all one really needs to do is switch to a different set of masks to make processors at different nm sizes. When one says that they have a 65nm process, that means that one has the masks and other gear that is accurate down to 65nm. That's accurate to 65nm. Similar to the concept that a number that is accurate to 100 decimal places is also accurate to 10 decimal places, the 65nm process can make anything less detailed, or larger, than 65nm.

Again, this isn't a hard concept...


--


Again, more posts from people who want to pretend like they know something. AMD has 65nm gear right now, and is making working silicon on it. AMD is selling 90nm CPUs off this line right now, and will transition the equptment they already have to 65nm as soon as they decide they can and they need to.

--

Let me say this for the benifit of everyone:

If you honestly know what you are talking about, post away. If you don't know what the hell you are talking about, don't pretend to. You know what that's called? Honesty! I swear, it feels like I'm in a highschool crowd full of little teens trying to bullshit all over the place to impress people: I find it horribly pathetic and depressing.

One of these days I'm going to leave all the "[H]ard" kids to bullshit among themselves and just give up on this forum.
Click to expand...

Duvie go back to ATF that kind of talk is not welcome in these forums. Go back to ATF and spread your FUD
 
Scali said:
While visaris' story is right from a technical point-of-view (65 nm equipment can build chips at pretty much any scale, down to 65 nm), technically AMD does not have a complete 65 nm process yet.
That is, they have 65 nm equipment, but either they don't have all parts of the production-line ready for 65 nm yet, or they are having problems with scaling their products down to 65 nm... because fact is, AMD is not producing 65 nm processors yet.
And it's not a good thing to invest in a 65 nm production-line, and continue to produce 90 nm chips, because you won't get a return on your investment.
I think AMD is going to have a very rough time when Conroe comes out. In fact, they may already be having a rough time, but it's not that apparent to the public when their chips are still performing well.
For Intel it was the other way around, lately. Intel had everything together with the 65 nm process, and their sales figures and profits were fine aswell... But the focus was on the 'bad' performance of the P4.
Click to expand...

You know, after this turns into the very first real major and pointless flame war on this forum and once the semantics and pendantics have died down, someone is going to look back at your post and quote it for truth. I applaud its the accuracy. Good job!
 
well i dont wanna create a war and i know im on intel turf...

and i really do swing for both teams, have two gaming comps and really there exact rivals to each other

but maybe we should compare intels conroe thats going to be released no shorter then several months, to something that amd will release no shorter then in several months? just a thought. and am2 i think is gonna give more of a boost then 5%. i mean it just has more pins. what can intel do about that?

even so if the conroe really does wipe the floor with amd (and im gonna wait till both am2 and conroe are out and only trust those benchies) then i will swap the mobo and processers of both these comps for one that supports conroe, and a conroe system!
 
Take a look at the past couple of posts. While I'm not an english major, and it's not my job to write a "Chip Fab 101" document for you, I explained several basic concepts pretty clearly. All you people posted as a response was:


$BangforThe$ said:
Duvie go back to ATF that kind of talk is not welcome in these forums. Go back to ATF and spread your FUD
Click to expand...
1c3d0g said:
Spare your breath, savantu. There's never enough ammo to fight trolls like visaris, unfortunately...
Click to expand...
savantu said:
You Sir , are one big idiot. Not only that you don't have a clue on chipmaking , but worse you jump on others who guess what , work at Intel , and they have at least an ideea of what chipmaking implies. The other problem is , you are so gullible that it hurts. (...) You can apologize to empoy if you are as "honest" as you say...but IMO , far from it.
Click to expand...

Where is your rebuttal to my argument that 65nm fabs can make 90nm chips? Where is your explanation of how I am wrong? Where do you outline the flaw in my line of thinking or logic? ... The answer is that you haven't done any of these things. You just like to sling insults around.

I don't work for AMD, so I can't prove what gear they have or don't have in their fabs. All I can do is post a link from a respected chip fabrication site, which I have done. Either you trust FabTech or you don't. I do, but you don't have to. We could argue all day about this issue and get nowhere.

--

That's not what really bothers me though. The fact that so many people keep fighting me on the fact that 65nm fabs can't make 90nm chips is what is distrubing and makes me question your knowledge and character. While I give explanations to the best of my ability, you give nothing of value, only insults and poorly thought out statements.

This is the fourth time in a row I'm going to explain this. Perhaps if I repeat it enough, it will sink in:

Many people here seem to think that a 65nm process creates chips where all the features are the same size at 65nm large. This is not the case. A process is described by the smallest feature it can produce. Some people claim the process is actually described by the average feature size, but this difference is totally irrelevent to my claim that 65nm gear can make 90nm chips. Here, let me do a google search for you and quote a few pages, because you can't seem to handle a search engine:

http://www.dbanks.demon.co.uk/ueng/minfeat.html said:
The minimum feature size is the width of the smallest line or gap that appears in your design.
Click to expand...
http://www.facsnet.org/tools/sci_tech/tech/fundaments/scale2.php3 said:
Chips are generally categorized by their "minimum feature size" -- the dimension of the smallest feature actually constructed in the manufacturing process.
Click to expand...
http://en.wikipedia.org/wiki/90_nanometer said:
The 90 nanometer process refers to the average feature size. However, the minimum feature size on 90 nanometer chips can actually be quite smaller, down to around 45 nanometers.
Click to expand...
http://www.csee.umbc.edu/~plusquel/611/slides/intro1.html said:
Feature size: Size of the smallest features on an IC, usually the length of the transistor channel.
Click to expand...

The point is that many different features on the CPU are different sizes, and the process technology (i.e. 65nm) typically refers to the smallest feature a manufacturing process can make. There is no reason that 65nm gear could not make a 90nm CPU. As long as one has the propper maks, there should be no issue whatsoever. Are you just confused about the meaning of the word "smallest"? You do realize all these quotes I have just made directly imply that anything larger than 65nm should be possible in a 65nm fab, don't you?

Again, if you don't think AMD has 65nm gear, and you don't want to trust my FabTech link, fine. Don't trust the link. But that is a totally separate issue from what a 65nm fab can and can't make.

So, again, I have done my best to explain how these things work. I think many of the people fighting me on this issue just can't get past the fact that the numbers 90 and 65 are not equal. If you have any depth to your argument, or see any flaw in my logic or thinking, I encurage you to post a response. I would love to have real discussion with someone about this issue.

If you can't come up with a counter-argument that has any depth to it, please just shut up. The shallow (I should say non-existant) arguments you make waste everyone's time, and make you seem like some highschool kid trying to bullshit in front of a crowd to look cool.
 
Thank you for explaining that.. I was merely trying to use what i know about making things.. (not very much...) and put it together with what very little i know about chip fabrication processes.. and i was having the hardest time finding anything out on google.. im not a very good searcher when i dont really know what im looking for.


wow.. a coherient post for once... maybe... :)
 
visaris said:
Where is your rebuttal to my argument that 65nm fabs can make 90nm chips? Where is your explanation of how I am wrong? Where do you outline the flaw in my line of thinking or logic? ... The answer is that you haven't done any of these things. You just like to sling insults around.
Click to expand...

A 65nm FAB capable that makes 90nm is a 90nm FAB.You can't produce both 65 and 90nm chips at the same time on the same production line.

A 65nm FAB has its tools , masks , layout technics and procedures designed for 65nm operation.

It's like writing in Word with Font size 12 vs 14.You either write with 12 or 14 , and changing size means doing an extra operation ( selecting the new size ).Just like that a FAB has to change its tools and modus-operandi if it moves from one feature size to another.

Many people here seem to think that a 65nm process creates chips where all the features are the same size at 65nm large. This is not the case. A process is described by the smallest feature it can produce. Some people claim the process is actually described by the average feature size, but this difference is totally irrelevent to my claim that 65nm gear can make 90nm chips. Here, let me do a google search for you and quote a few pages, because you can't seem to handle a search engine:
Click to expand...


The smallest feature size it can produce doesn't describe the process.Intel's 90nm process had its smallest feature size 50nm while , its 65nm has 35nm.AMDs 65nm process operates the smallest feature at 30nm.

The point is that 65nm gear which produces 90nm chips isn't "65nm" gear.It is 90nm gear.End of story.That gear is modified from 90nm to 65 and you fail to see this.

Secondly , the link I gave you said explicitly that 65nm wasn't installed in FAB 36 , but AMD was borrowing money to buy it.Hector Ruiz , AMD CEO , said the same : we are producing 90nm chips and we'll prepare for 65nm in 2006".

When you'll be able to get past corporate spin and idiotic preconceived ideas come back...
 
savantu said:
A 65nm FAB capable that makes 90nm is a 90nm FAB.You can't produce both 65 and 90nm chips at the same time on the same production line.
Click to expand...

so what machines cant change bits now huh, guess that means my wood working project is fucked.
 
MrWizard6600 said:
so what machines cant change bits now huh, guess that means my wood working project is fucked.
Click to expand...

Changing process size in a FAB is a little bit more complicated than a wood working project....Buy hey , some people are full of surprises!
;)
 
http://www.theinquirer.net/?article=30385

AMD IS GEARING up its latest baby, its Fab 36 factory based in Dresden, Germany. I visited Dresden two weeks ago, which is when the ramp up began. The first 12-inch wafers are already in production. While the reps declined to comment what processors are being printed on the wafers, the press release is expected quite soon. The company execs recently gave the green light to ramp up the production to 13.000 wafer starts a week. How many of those wafers are usable, remains to be seen.
However, the situation is not entirely rosy. We've heard partners and distributors spreading a lot of bitter comments regarding AMD's tight supply of the processors, which is allegedly due to the "extreme success" of the Opteron Series 800's, which customers are really buying like there's no tomorrow. AMD currently has 40 per cent of the world's 4P server market, and the company reps are confident the firm will take the lead in that segment before the end of the year.

The other buzz is that AMD has yield problem, and that's the main reason why some wafers are being built with strained SOI and others on "regular" SOI wafers.

Fab 36 should help re-solve the supply story in any case, but the equipment has not yet arrived in full force. One assembly line is completed, there are several more to follow. Bear in mind that both Fab 30 and 36 aren't built using the normal "pipeline" concept, but rather with an overlapping one. The 65 nanometre assembly kit is also ordered, and AMD expects to start printing 65 nm lithography in the next calendar quarter, ready for a Q4 introduction. µ
Click to expand...

no 65nm "gears" yet :rolleyes:
 
ROFL!

You guys really crack me up. So far you still haven't posted anything except "Teh numbers 90nm and 65nm are not the same, haha! I wonder if he'll say he's sorry!"

You just go ahead be wrong :) I really am over trying to educate you ;)
 
A 65nm FAB capable that makes 90nm is a 90nm FAB.You can't produce both 65 and 90nm chips at the same time on the same production line.
Click to expand...

Seems you dont follow history very well....

AMD had transitioned to 130nm while still in production with 180nm chips, they did the same thing again when they transitioned to 90nm..... ALL on fab30

It's like writing in Word with Font size 12 vs 14.You either write with 12 or 14 , and changing size means doing an extra operation ( selecting the new size ).Just like that a FAB has to change its tools and modus-operandi if it moves from one feature size to another.
Click to expand...

Seems you dont know jack about process technology either.... The comparison you are trying ot make... doesnt make any bit of sense.....

A better example would be a laser printer (bare with me) It can choose to print on whatever resolution the printer supports. If you think about it that way.... That is more or less what lithography is... (although MUCH more complex) Set your desired resolution and print....

The point is that 65nm gear which produces 90nm chips isn't "65nm" gear.It is 90nm gear.End of story.That gear is modified from 90nm to 65 and you fail to see this.
Click to expand...

Ummm Isnt that why it is called a transition? It doesnt matter what equiipment is installed into the facility... The FACILITY was designed to handle 65nm production..... The equipment comes later to make the FACILITY work.... Get it?

When you'll be able to get past corporate spin and idiotic preconceived ideas come back...
Click to expand...

What are you talkin about? AMD announced along time ago that they would begin fab36's ramp up on 90nm.... The reason they gave was to analyze the FACILITIES yeild curve, then when they get yeilds to an acceptable level they will begin the 65nm ramp up....

They did the same thing on fab30 when they opened it... Remember? They did the initial ramp up on 250nm then later transitioned to 180nm...

I guess all this discussion did was show your lack of knowledge. It seems quite clear that you dont bother to learn the history of the market that you claim knowledge in, and that you also dont study the market that you claim knowledge in.....

Last I heard AMD's fab36 90nm process is supposed to about 30% faster than fab30's 90nm process.... That means that AMD will be able to release 90nm chips based on rev F at or around 3.2-3.4 ghz

Also keep in mind that AMD's 65nm process supports FDSSOI (Fully Depleted Strained Silicon On Insulator) Which will allow it's 65nm process to scale alot higher than that.... Couple that with K8L next year..... (about the time when Conroes paper launch will finally be close to an end, when you can actually buy the chip)
 
Last I heard AMD's fab36 90nm process is supposed to about 30% faster than fab30's 90nm process.... That means that AMD will be able to release 90nm chips based on rev F at or around 3.2-3.4 ghz
Click to expand...

haven't seen any any 3.2-2.4Ghz Athlon64 cpus. If Fab36 is operational already and is processing 90nm chips we should have seen them already.

Also keep in mind that AMD's 65nm process supports FDSSOI (Fully Depleted Strained Silicon On Insulator) Which will allow it's 65nm process to scale alot higher than that.... Couple that with K8L next year..... (about the time when Conroes paper launch will finally be close to an end, when you can actually buy the chip)
Click to expand...

again, no 65nm "gears" inside Fab36 :rolleyes:
 
empoy said:
haven't seen any any 3.2-2.4Ghz Athlon64 cpus. If Fab36 is operational already and is processing 90nm chips we should have seen them already.
Click to expand...
who said that fab36 is really putting out much already? i thought everything said thus far was "ramping up"
plus, i've seen quite a few fx-57's that easily do 3.2-3.4ghz.. :p
 
AFAIK All of the chips being produced at fab36 is the ramp up for rev F (AM2 S1 F) S o yeah they should be in production, but you wont see them untill the new sockets get released.

again, no 65nm "gears" inside Fab36
Click to expand...

Like I said it is the Facility that matters most. I'm no guru on fabs, but It has alot to do with oddball things like how much air the purifying system can circulate. How much electrical capacity the system has, and how clean the supply is. And a bunch of other stuff that I couldnt even begin to explain....

That is the reason why it would be prohibitive for AMD to transition fab30 to 65nm becouse they would have to overhaul the facility.... It just doesnt meet the requiremnts thta 65nm production demands.

Rumors that I've heard say that it was amazing that AMD even managed to pull off 90nm... You really got to give them credit. They have received more awards in fabrication than ANY other semiconductor company in the world. They have the most advanced process technology in the world right next to IBM.... Some apects are somewhat outdated, for example they are still using a somewhat small wafer at fab30, but they still get fab of the year every year inspite of that.
 
who said that fab36 is really putting out much already? i thought everything said thus far was "ramping up"
Click to expand...

a couple of your comrads have said that fab36 is operational and I quote "The devices being fabricated at Fab36 use AMD's 65nm process but on 90nm design rules to assist in a fast ramp with high yields." :D

Fab36 is supposed to start ramping-up to 65nm but I guess AMD doesn't have money to buy them as of now :)

plus, i've seen quite a few fx-57's that easily do 3.2-3.4ghz
Click to expand...
but these were fused below 3Ghz. on that fact alone, AMD is not confident that it'll be stable on 3.2-3.4Ghz.
 
You must log in or register to reply here.
Back
Top