Nvidia confirms the GTX 580

[Iconyu]

Arrgh... Sam Oslo sounds like a broken record. Is there any way to put him to ignore list or something, because he isn't contributing anything useful information any of these GFX threads? Endless debate about new node being new architecture, yada yada yada...

Click on his name and then click on view profile, click on user lists and the option is right there.

[FischOderAal]

Is there any way to put him to ignore list or something, because he isn't contributing anything useful information any of these GFX threads?

http://www.xtremesystems.org/forums/...?do=ignorelist

You're welcome.

Click on his name and then click on view profile, click on user lists and the option is right there.

Your way is even easier.

[Jowy Atreides]

I've said already, a shrink is the least you need to call it new generation (and change the first digit in number-game).

Tick tock.


New Microarchitecture.
New Process Technology.
New Microarchitecture.
New Process Technology.
New Microarchitecture.
New Process Technology.

Separate things.

In computer engineering, microarchitecture (sometimes abbreviated to µarch or uarch), also called computer organization, is the way a given instruction set architecture (ISA) is implemented on a processor. A given ISA may be implemented with different microarchitectures.[1] Implementations might vary due to different goals of a given design or due to shifts in technology.[2] Computer architecture is the combination of microarchitecture and instruction set design.

/argument.

Unless you think you have far more information than intel and they're wrong(?)

[Sam_oslo]

Can you PLEASE educate yourself by browsing the web & look for info regarding nVidia G71 and nVidia G80, both built using THE SAME TSMC 90 nm process, or ATi R200 and ATi R300, both built using THE SAME TSMC 150 nm process ???

If you don't know jack about GPU mArch and won't learn even one bit, please stay away from a discussion regarding a more sophisticated topic like GPU generational change, your stubborness & ignorance is just so annoying.



Echoing your frustatic scream from a tropical archipelagic nation down under, aarrrggghh ....

Learn yourself to respect others ideas, before preaching education to others.

This is my ideas, we need a shrink to call it a new generation, otherwise we are going to get the same mess and confusions we had win G92-mess in this round form both AMD and nVidia. Just wait and watch.

Those who don't like my idea are free to express it in a constructive way. Those who are trying to get aggressive with personal attacks, I gotta say, just eat it.

[Jowy Atreides]

Learn yourself to respect others ideas, before preaching education to others.

Peoples ideas can be outright incorrect and respecting them would be disrespecting logic and intelligence. There comes a point when an Idea is outright proven wrong and is discarded.
Have one of my ideas: Gravity occurs as a result of tiny sub-atomic Ralph Macchios waxing on and waxing off nearby atoms, pushing them towards the Planets surface.

I'll now proceed to tell you that you are wrong for 4 pages then say you should respect my ideas even though I clearly don't respect yours.

To end this, I will proclaim;
IT IS RAPLH MACCHIO AND HE IS ANGRY YOU FOOL!!!

Your idea is not an idea any longer, it has been proven incorrect and is now classed as a lie or untruth.

[Xoulz]

Learn yourself to respect others ideas, before preaching education to others.

This is my ideas, we need a shrink to call it a new generation, otherwise we are going to get the same mess and confusions we had win G92-mess in this round form both AMD and nVidia. Just wait and watch.

Those who don't like my idea are free to express it in a constructive way. Those who are trying to get aggressive with personal attacks, I gotta say, just eat it.

Ur idea doesn't constitute fact..

[Iconyu]

http://www.xtremesystems.org/forums/...?do=ignorelist

You're welcome.

Your way is even easier.

Thanks, I've been planning to use it for a while and his last post drove me over the edge. He was just too stupid to tolerate.

[Sam_oslo]

Peoples ideas can be outright incorrect and respecting them would be disrespecting logic and intelligence. There comes a point when an Idea is outright proven wrong and is discarded.
Have one of my ideas: Gravity occurs as a result of tiny sub-atomic Ralph Macchios waxing on and waxing off nearby atoms, pushing them towards the Planets surface.

I'll now proceed to tell you that you are wrong for 4 pages then say you should respect my ideas even though I clearly don't respect yours.

To end this, I will proclaim;
IT IS RAPLH MACCHIO AND HE IS ANGRY YOU FOOL!!!

Your idea is not an idea any longer, it has been proven incorrect and is now classed as a lie or untruth.

You have the right to dismiss my ideas and defend yours. But you can't insist on having patent on ultimate truth, and be the judge at the same time too.

A few are preaching to be the judge of this, so tell me, if the shrink is not the norm, where does the line go for generation-shift? A better performance/power usage would do? How about a re-fresh or a re-brand?

[Solus Corvus]

Learn yourself to respect others ideas, before preaching education to others.

This is my ideas, we need a shrink to call it a new generation, otherwise we are going to get the same mess and confusions we had win G92-mess in this round form both AMD and nVidia. Just wait and watch.

Those who don't like my idea are free to express it in a constructive way. Those who are trying to get aggressive with personal attacks, I gotta say, just eat it.

These guys are being rude about it, but they are right. Architecture and process are two separate things. Architecture is the number and type of units in the core, their arrangement and connections, and other such details. Process is the physical implementation of an architecture onto a chip. An architecture could be implemented on different processes. And a particular process could be used to produce very different architectures. But a new generation requires a new architecture, not a new process as the examples already presented to you show. Are you saying G71 to G80 wasn't a new generation?

Saying that we should call something on the same process a new generation because it will confuse people is a logical fallacy called an Appeal to Consequences. Words have meaning, even if we don't like the results of that meaning.

[Xoulz]

nvm

[Sam_oslo]

These guys are being rude about it, but they are right. Architecture and process are two separate things. Architecture is the number and type of units in the core, their arrangement and connections, and other such details. Process is the physical implementation of an architecture onto a chip. An architecture could be implemented on different processes. And a particular process could be used to produce very different architectures. But a new generation requires a new architecture, not a new process as the examples already presented to you show. Are you saying G71 to G80 wasn't a new generation?

Saying that we should call something on the same process a new generation because it will confuse people is a logical fallacy called an Appeal to Consequences. Words have meaning, even if we don't like the results of that meaning.

We don't know what kind of architectural changes we are going to get with upcoming high-end GPUs. I'm afraid we are going to see the same mess as G92 in this round, form both AMD and nVidia. (just as a speculation for now, but just wait and watch).

I've been trying to raise the question, where does the line go for a architectural changes to become a new generation? But nobody has given a clear answer. Could you cast some light on it?

Architectural-changes is a trickery term, and when I look at 68xx, it's smaller, cooler and cheaper, but is it a new generation? In case why does it perform less that the last gen with exact same name?
We must have a clear definition to define a new generation (with new number at first digit), and I'm open to see what else could be used as a clear definition, but these kids are taking it personal.


It's the nature of these GPU-treads, they have tendency to get, funny, rude, aggressive, and finally to childish fights. But it doesn't scare me, those who are trying can just eat it.

[Solus Corvus]

I've been trying to raise the question, where does the line go for a architectural changes to become a new generation? But nobody has given a clear answer. Could you cast some light on it?

No, there is no hard line. Architectures are so complex that sometimes a small change in one part could result in a major improvement in performance, while major changes to another part might only make minimal difference. There isn't going to be any such hard definition and often it just depends on what the company wants to call a new generation.

Personally I would call 68xx an evolution of the evergreen generation, and did so shortly before launch based on the rumors. We will have to see about Cayman. But as we know, in computers as in biology, evolution is driven by small changes in successive generations (to twist the analogy a bit) .

[Boissez]

We don't know what kind of architectural changes we are going to get with upcoming high-end GPUs. I'm afraid we are going to see the same mess as G92 in this round, form both AMD and nVidia. (just as a speculation for now, but just wait and watch).

I've been trying to raise the question, where does the line go for a architectural changes to become a new generation? But nobody has given a clear answer. Could you cast some light on it?

Architectural-changes is a trickery term, and when I look at 68xx, it's smaller, cooler and cheaper, but is it a new generation? In case why does it perform less that the last gen with exact same name?
We must have a clear definition to define a new generation (with new number at first digit), and I'm open to see what else could be used as a clear definition, but these kids are taking it personal.


It's the nature of these GPU-treads, they have tendency to get, funny, rude, aggressive, and finally to childish fights. But it doesn't scare me, those who are trying can just eat it.

You are way to hung up on naming. Marketing is, by the end of the day, who gets to name products - not R&D.

And whether a new product is a new generation or not is semantics and can be discussed ad nauseum.

[Sam_oslo]

No, there is no hard line. Architectures are so complex that sometimes a small change in one part could result in a major improvement in performance, while major changes to another part might only make minimal difference. There isn't going to be any such hard definition and often it just depends on what the company wants to call a new generation.

Personally I would call 68xx an evolution of the evergreen generation, and did so shortly before launch based on the rumors. We will have to see about Cayman. But as we know, in computers as in biology, evolution is driven by small changes in successive generations (to twist the analogy a bit) .

I agree with most part of you comment, specially when you say 68xx is not new generation. Exactly as GF104 (GTX 460) was not a new generation, even tho it was smaller, cooler, and cheaper.

But we need to make awareness, by raising this question, and find a clear definition. Otherwise, these guys (AMD and nVidia both of them) are going to call these gray-zone evolution/improvements as new generation, and soon run out of new numbers too, LOL.

As you said, the details about upcoming high-end GPUs is still unclear, but I'm afraid they are going to mess it up just for the purpose of marketing (as other are suggesting over too). Whatever is behind it, it will make a lot of confusion, but I want to wait until I see them before jumping to conclusion.

[foch3 -USA-]

Sam please drop it. I am sure you thought you had an interesting idea, and it made sense to you somehow but it's just derailing threads.

Quite interesting that nvidia has lowered the price of fermi quadro cards so drastically. Either GTX580 is a big step or they're scared of cayman based professional cards.

[Sam_oslo]

Sam please drop it. I am sure you thought you had an interesting idea, and it made sense to you somehow but it's just derailing threads.

Quite interesting that nvidia has lowered the price of fermi quadro cards so drastically. Either GTX580 is a big step or they're scared of cayman.

Is it bordering you? I'm taking about upcoming high-end GPUs, and trying to find out if they are "fake" or "real" generation. Maybe I've repeated some, in case, sowwy but I felt the need to raise awareness.

I would love to talk about the details about GTX580 and Cayman, it would be much more interesting, but I don't have much reliable info yet. Just drop what you got, you won't disturb, and I will follow you.

[Solus Corvus]

I agree with most part of you comment, specially when you say 68xx is not new generation. Exactly as GF104 (GTX 460) was not a new generation, even tho it was smaller, cooler, and cheaper.

But we need to make awareness, by raising this question, and find a clear definition. Otherwise, these guys (AMD and nVidia both of them) are going to call these gray-zone evolution/improvements as new generation, and soon run out of new numbers too, LOL.

As you said, the details about upcoming high-end GPUs is still unclear, but I'm afraid they are going to mess it up just for the purpose of marketing (as other are suggesting over too). Whatever is behind it, it will make a lot of confusion, but I want to wait until I see them before jumping to conclusion.

Well I agree that companies calling a barely changed architecture a new generation is marketing driven. But just as we are able to say that "I don't think this card should be called X" we could just as easily say that we don't think certain chips should be considered new generations.

But I don't think that requires us to make up some specific definition for generation just to avoid undesirable consequences. If we could even agree on a definition at all. Instead I think we should simply be open to discussing what a company calls a generation and deciding for ourselves if it really deserves that moniker. And I don't think we should be overemphasizing the importance of defining something as fitting in a specific generation when the the actual performance of a product is so much more complex then what can be defined by a single word.

[Chumbucket843]

These guys are being rude about it, but they are right. Architecture and process are two separate things. Architecture is the number and type of units in the core, their arrangement and connections, and other such details. Process is the physical implementation of an architecture onto a chip. An architecture could be implemented on different processes. And a particular process could be used to produce very different architectures. But a new generation requires a new architecture, not a new process as the examples already presented to you show. Are you saying G71 to G80 wasn't a new generation?

Saying that we should call something on the same process a new generation because it will confuse people is a logical fallacy called an Appeal to Consequences. Words have meaning, even if we don't like the results of that meaning.

maybe it's hard for me to separate physical and architecture differences because my perspective is more design oriented but the two are related. for example, arch A may create a nasty physical design while arch B has a good one, but arch B might have some crippling architectural bottleneck. arch B could be prescott in the P4 or arch A could be fermi in GF100.

in my opinion a next generation gpu can be defined roughly as a gpu that is capable of something previously not possible and in a meaningful way (5% faster wont cut it.). fitting barts into this is difficult, but with 20/20 hindsight it will become clear if it's a next gen gpu.

also i have something important to note about processes. less nanometers does not make a process better. arguing that 28nm is faster than 40nm because it has smaller structures is not true. physics gets very strange at this level and is very counterintuitive. this means that a new process does not make or break a next generation. do you like how i keep my sentences all the same length? i like being monotonous and terse.

[Solus Corvus]

Chumbucket, I agree. I was (over?)simplifying a lot.

[Calmatory]

Is it bordering you? I'm taking about upcoming high-end GPUs, and trying to find out if they are "fake" or "real" generation. Maybe I've repeated some, in case, sowwy but I felt the need to raise awareness.

I would love to talk about the details about GTX580 and Cayman, it would be much more interesting, but I don't have much reliable info yet. Just drop what you got, you won't disturb, and I will follow you.

It is up to the manufacturer to decide whether or not a product is of a new generation. It is up to you to agree or disagree with it, however, you aren't the one to decide whats right or wrong.

If AMD calls Cayman 7xxx and claims its a new gen stuff, then it is. You can disagree with it, but thats all. They can decide, you just have an opinion.

[foch3 -USA-]

Is it bordering you? I'm taking about upcoming high-end GPUs, and trying to find out if they are "fake" or "real" generation. Maybe I've repeated some, in case, sowwy but I felt the need to raise awareness.

I would love to talk about the details about GTX580 and Cayman, it would be much more interesting, but I don't have much reliable info yet. Just drop what you got, you won't disturb, and I will follow you.

We know next to nothing about them as you said yet your worried if their new generations? Geforce 5200 was slower than geforce 4 4200 even though it was on a new process and a different architecture. Faster is not always a necessity for a new generation. I would consider gf104 to be a new generation because of the reordering of shader's and its proximity in performance compared to the larger gf100.

[Sn0wm@n]

Learn yourself to respect others ideas, before preaching education to others.

This is my ideas, we need a shrink to call it a new generation, otherwise we are going to get the same mess and confusions we had win G92-mess in this round form both AMD and nVidia. Just wait and watch.

Those who don't like my idea are free to express it in a constructive way. Those who are trying to get aggressive with personal attacks, I gotta say, just eat it.

LOL a shrink isnt a new arch .... deffinitions have been posted yet you go back to your logic ....

Peoples ideas can be outright incorrect and respecting them would be disrespecting logic and intelligence. There comes a point when an Idea is outright proven wrong and is discarded.
Have one of my ideas: Gravity occurs as a result of tiny sub-atomic Ralph Macchios waxing on and waxing off nearby atoms, pushing them towards the Planets surface.

I'll now proceed to tell you that you are wrong for 4 pages then say you should respect my ideas even though I clearly don't respect yours.

To end this, I will proclaim;
IT IS RAPLH MACCHIO AND HE IS ANGRY YOU FOOL!!!

Your idea is not an idea any longer, it has been proven incorrect and is now classed as a lie or untruth.

LOL so much sig material in this post ... wow ... wax on wax off you stupid gravity

[tifosi]

I agree with Diltech in basically this 5 series name is being forced by AMD hand. It would be harder to sell a 4 series when AMD has released released its 6 series parts.

If we look at the juniper to barts and gf100--> gf104 technology, their is probably more differences in the architecture of gf104 than barts shift.


If AMD didn't change their series to 6xxx, Nv probably wouldn't have gone with a gtx 5xx name change.

Man... i was occupied with work and i was worried sick about task i'm to perform... but then you come and... I want whatever you're on!

Nvidia screwed up badly... just because your competitor is better, doesn't mean you have an automatic right to deceive people left, right and center. Jensen is too egotistical, and some Nvidia fans, just wow!!!

AMD's 6 series cards which we see (Barts) they're supposed to be replacement for 5770 and lower cards... They're from a new generation. It's like moving from Nehalem to Sandy Bridge from Intel.

[cold2010]

GTX 580

[tajoh111]

Man... i was occupied with work and i was worried sick about task i'm to perform... but then you come and... I want whatever you're on!

Nvidia screwed up badly... just because your competitor is better, doesn't mean you have an automatic right to deceive people left, right and center. Jensen is too egotistical, and some Nvidia fans, just wow!!!

AMD's 6 series cards which we see (Barts) they're supposed to be replacement for 5770 and lower cards... They're from a new generation. It's like moving from Nehalem to Sandy Bridge from Intel.

What with you guys trying to antagonize other people and start flaming wars?

This new generation of fans is really making this forum a worse place.

Anyways, it doesn't give you the right but marketing wise it makes far more sense. If NV kept the gtx 4xx moniker, they would likely sell less cards because of it. This is because a lot of people put weight in naming, they would be thinking its still last generations card which is half true. This is an act of desperation for a company that has fallen behind in the development cycle when it comes to performance per watt(they still compete quite well for performance). This naming scheme has nothing to do with Jensen and everything to do with market conditions and a flawed product.

None the less your getting more performance for likely the same or less money, which in turn helps lessen the blow of AMD new naming scheme and products. Its kind of a move for AMD to shift its naming upward, just as it was a move for NV to rebrand cards.

Since your all about criticizing naming schemes by consequence of what your saying against Nvidia for trying to deceive it customers, what do you think of AMD latest naming tactics. When your beating up your competitors at selling cards, why start shifting your naming scheme so it is more confusing for customers, when you already have the advantage? Nvidia is doing part of this out of desperation and because the market demands it. Why is AMD doing it?

AMD is just doing it so people swallow the barts price easier because most people would have a hard time paying 239 and 179, when last generations 577x series were 159 and 129. It's dishonest because never in the history of ATI videocards has buying a card in the same series gotten you lower performance(besides when a naming scheme changes entirely). Still even if AMD didn't drop the price and kept barts xt priced as it is, I think all of us would have appreciated the naming scheme to stay 67xx because it is more honest. Barts codename was rv940 which for the longest time was AMD midrange. If we accept barts xt as a next generation midrange, its performance is inline where it should have been, but its price is significantly higher than people thought, as alot of people were thinking barts xt was going to be under 200 dollars.

Both naming acts are dishonest, but a lot more people are suffering from AMD naming scheming because if you do your research and all that jazz, you will likely to be able to sidestep both problems and with both naming schemes. However by pricing Barts XT for what it is, people that want to buy a card at around 150-100 are SOL and have to settle for last generations stuff as roadmaps indicate that the 57xx gen not being discontinued. AMD going to make more money, but this probably doesn't benefit the consumer.