Foundry Delays to Push Back NVIDIA Kepler Launch to 2012

[Bjorn_Of_Iceland]

Good.. that gives me one more year to save up before changing the GTX580

 

[Benetanegia]

To be honest, what it looks like to me is that NVidia is more popular with "some" Steam users surely? That don't mean they are selling more cards though

Explain what you mean please. The survey takes ALL Steam users into account.

I hope that you are not suggesting that there's any kind of preference for Nvidia amongst Steam users, and that for non Steam users AMD is prefered. That would be ridiculous.

For me the survey is enough proof, but sure is not the only one:

http://www.jonpeddie.com/publications/add-in-board-report/

That Nvidia sells more cards is a fact. That Nvidia tends to dominate the higher end segments, while AMD sells more on the lower end segments (and integrated), is also suggested by nearly every analist. Since we are talking about Kepler, hence high-end, I could care less how many cards are sold in the low end segments (mostly for HTPC) and even taking all segments into account Nvidia sells 60% of cards. Pretending that reality is different, because it does not relate to your graphics card election is wrong. Don't try to find some obscure reasons for explaining what is cristal clear by just looking at facts...

 

[Tatty_Two]

Explain what you mean please. The survey takes ALL Steam users into account.

I hope that you are not suggesting that there's any kind of preference for Nvidia amongst Steam users, and that for non Steam users AMD is prefered. That would be ridiculous.

For me the survey is enough proof, but sure is not the only one:

http://www.jonpeddie.com/publications/add-in-board-report/

That Nvidia sells more cards is a fact. That Nvidia tends to dominate the higher end segments, while AMD sells more on the lower end segments (and integrated), is also suggested by nearly every analist. Since we are talking about Kepler, hence high-end, I could care less how many cards are sold in the low end segments (mostly for HTPC) and even taking all segments into account Nvidia sells 60% of cards. Pretending that reality is different, because it does not relate to your graphics card election is wrong. Don't try to find some obscure reasons for explaining what is cristal clear by just looking at facts...

You precisely stated my point, thank you, it's steam users and whilsts steam is huge, it's by no means all users and therefore, whilst it's an excellent measure of card popularity for the many that use steam (and have either participated in the survey or bothered to update their specs if it's automatic) it does not reflect all users.

For example, If I go to my account in Steam I think it shows that I have a GTX280, thats only because I havent bothered logging on to steam for about 2 years..... nothing more sinister than that, so I can confirm that your conclusion that I might be insinuating that steam is biased in favour of NVidia is unfounded...... clearer?

Edit: noone.... well I am not and have not suggested that anyone sells more cards than NVidia, I am simply pointing out that in my opinion Steam does not reflect all users and there is not "firm" evidence, thats not to say I think your wrong, meerly that there may be more steadfast ways of prooving your point.

 

[Benetanegia]

You precisely stated my point, thank you, it's steam users and whilsts steam is huge, it's by no means all users and therefore, whilst it's an excellent measure of card popularity for the many that use steam (and have either participated in the survey or bothered to update their specs if it's automatic) it does not reflect all users.

For example, If I go to my account in Steam I think it shows that I have a GTX280, thats only because I havent bothered logging on to steam for about 2 years..... nothing more sinister than that, so I can confirm that your conclusion that I might be insinuating that steam is biased in favour of NVidia is unfounded...... clearer?

I know that Steam does not represent the entirety of users (no survey does), but there's no better survey out there. You seem to suggest that the very first thing that every Nvidia user does is enter Steam while AMD users never use Steam. That is ridiculous like I said.

And if there is any preference, that would be for AMD, because Valve has been promoting Ati cards on their games for a very long time. I'm not saying there's any preference, only that if there had to be one, it makes more sense to be in favor of Ati. Of the 5 Ati cards that I've bought in the last years (for me or friends) 3 had a discount cupon for some Steam game, none of the Nvidia ones had any.

And just as a side note, in 2010 50% of games sales were digital, and most of them are Steam. So Steam represents PC gamers better than anything else.

 

[Tatty_Two]

I know that Steam does not represent the entirety of users (no survey does), but there's no better survey out there. You seem to suggest that the very first thing that every Nvidia user does is enter Steam while AMD users never use Steam. That is ridiculous like I said.

And if there is any preference, that would be for AMD, because Valve has been promoting Ati cards on their games for a very long time. I'm not saying there's any preference, only that if there had to be one, it makes more sense to be in favor of Ati. Of the 5 Ati cards that I've bought in the last years (for me or friends) 3 had a discount cupon for some Steam game, none of the Nvidia ones had any.

And just as a side note, in 2010 50% of games sales were digital, and most of them are Steam. So Steam represents PC gamers better than anything else.

You seem to suggest that the very first thing that every Nvidia user does is enter Steam while AMD users never use Steam. That is ridiculous like I said.

Not sure where you get this from, we seem to speak the same language but you continue to make wrong assumptions.... So I will say it again, and hopefully it will be even clearer, I don't care who sells the most cards, i dont care if more or less Nvidia or AMD users register on steam, what I do care about however is to try and get the fullest and most accurate picture when stating facts, you have just said that 50% of games were digital in 2010, that means 50% were not, therefore just by the law of averages there are probably as many gamers out there that dont play on steam (actually probably a lot more), so as I said before, whilst steam is a good indicator, it clearly does not speak for all gamers..... nothing more sinister than that, so please, stop this assumptions rubbish, you don't have to turn everything into a green or red side or assume everyone else is.

 

[Benetanegia]

You seem to suggest that the very first thing that every Nvidia user does is enter Steam while AMD users never use Steam. That is ridiculous like I said.

Not sure where you get this from, we seem to speak the same language but you continue to make wrong assumptions.... So I will say it again, and hopefully it will be even clearer, I don't care who sells the most cards, i dont care if more or less Nvidia or AMD users register on steam, what I do care about however is to try and get the fullest and most accurate picture when stating facts, you have just said that 50% of games were digital in 2010, that means 50% were not, therefore just by the law of averages there are probably as many gamers out there that dont play on steam (actually probably a lot more), so as I said before, whilst steam is a good indicator, it clearly does not speak for all gamers..... nothing more sinister than that, so please, stop this assumptions rubbish, you don't have to turn everything into a green or red side or assume everyone else is.

I'm not making this into a red/green anything. I'm stating the fact that Nvidia sells more, because that's a fact, it's information. Information to take it as you please. Because it sells more does it meas it's better? No. I'm even trying to imply that? No. But facts are facts. The post I replied to was implying that AMD was selling tons and tons of HD6000 cards more than GTX500, so that they didn't need the new generation as much. I pointed out that they are not selling so many of them, and provided proof. Proof that GTX500 vs HD6000 seems to be leaning in favor of Nvidia, and hence AMD needs to do something to change that trend.

Now regarding the discussion between you and me. You think that Steam survey is not a good source. And you are wrong and I'm telling you why. 50% of games sold digital, most from Steam. By the law of average, and contrary to your assumption, FAR more than 50% of gamers must use Steam, because it only takes 1 game bought through steam to be an Steam user. If this year you bought 100 games and only 1 was from Steam, guess what? You are an Steam user.

And honestly, I still fail to see the relevance of your posts. It doesn't matter if Steam users represent 50%, or 5% of gamers*. To think that a higher proportion of Nvidia users than AMD users get Steam is ridiculous and nonsensical.

* How many serious surveys do you know that were conducted with more than 5% of the targeted population? I'll save your time: NONE.

 

[a_ump]

I'm not making this into a red/green anything. I'm stating the fact that Nvidia sells more, because that's a fact, it's information. Information to take it as you please. Because it sells more does it meas it's better? No. I'm even trying to imply that? No. But facts are facts. The post I replied to was implying that AMD was selling tons and tons of HD6000 cards more than GTX500, so that they didn't need the new generation as much. I pointed out that they are not selling so many of them, and provided proof. Proof that GTX500 vs HD6000 seems to be leaning in favor of Nvidia, and hence AMD needs to do something to change that trend.

Now regarding the discussion between you and me. You think that Steam survey is not a good source. And you are wrong and I'm telling you why. 50% of sold digital, most from Steam. By the law of average, and contrary to your assumption, FAR more than 50% of gamers must use Steam, because it only takes 1 game bought through steam to be an Steam user. If this year you bought 100 games and only 1 was from Steam, guess what? You are an Steam user.

And honestly, I still fail to see the relevance of your posts. It doesn't matter if Steam users represent 50%, or 5% of gamers*. To think that a higher proportion of Nvidia users than AMD users get Steam is ridiculous and nonsensical.

* How many serious surveys do you know that were conducted with more than 5% of the targeted population? I'll save your time: NONE.

Well said on that last part. I agree, yes Steam survey isn't 100% or even 90%, but i'd bet my computer its the closest information we have to the exact reality on gaming desktop hardware percentages. IF more accurate/better data can be found please do provide

 

[Velvet Wafer]

GloFo seriously needs to create more fabs, to increase their production capacity... would also definetly benefit our economy, and create new working places

 

[laszlo]

i think this is old news ref.the 28nm issues...

http://blog.shrinkingviolence.com/2010/02/TSMC-40nm-28nm-yield.html

 

[KashunatoR]

it's amd which should be in the hurry to realease the new series of cards since the gtx 580 is kicking their ass really badly. if you OC the gtx 580 is enough to play any game at highest settings (fps doesn't drop below 45 in crysis 1 maxxed out 1080P!!!). but you could always use more power. for me the SLI is not a desirable solution. i've tried SLI gtx 480 and gtx 580 and the scaling is nowhere near it is advertised. i'll always go for the fastest single gpu card and i have a haunch it will be again nvidia who will deliver just that while amd fanboys will say that their cards have better performance/watt ratio. like that matters on the top products...

 

[dj-electric]

it's amd which should be in the hurry to realease the new series of cards since the gtx 580 is kicking their ass really badly.

uhm.... no. you got a graphics card that consumes more, cost a lot more and well... preform better (not a lot better). what do you expect? there is no magic here.
judging by the prices on new egg and comparing to reviews by w1zz, you will get the cheapest GTX580 for 499$ and the cheapest HD6970 for 329$ (both reference),
price differences = 51%
performance differences (2560X1600) = 9%
Power Consumption differences (peak) = 22%
So... you should probably go with a pair of HD6870 for 350$ total. there are enough good graphics solutions below 499$

 

[Benetanegia]

uhm.... no. you got a graphics card that consumes more, cost a lot more and well... preform better (not a lot better). what do you expect? there is no magic here.

There is no magic, but it's faster and has no competing product. That's why it costs more, it's an unrivaled flagship product.

They don't need the 580 to compete with AMD's best, because they already have cards that do that. In every high-end and performance spot, Nvidia has cards that directly compete with AMD offerings and currently sell for $30 less or so.

GTX570 which is as fast as the HD6970 is selling for $30 less and consumes about the same. It consumes a little bit more when load is 100% and a little less than the 6970 when on idle or playing blu-ray.

Same for GTX560, way cheaper than 6950, similar performance, small consumption difference, smaller die-area... For every high-end/performance AMD card Nvidia has a good answer that is selling at better prices and nobody really cares for 20W more on power consumption.

judging by the prices on new egg and comparing to reviews by w1zz, you will get the cheapest GTX580 for 499$ and the cheapest HD6970 for 329$ (both reference),
price differences = 51%
performance differences (2560X1600) = 9%
Power Consumption differences (peak) = 22%
So... you should probably go with a pair of HD6870 for 350$ total. there are enough good graphics solutions below 499$

Yes and you can get a GTX570 for $309.

The fact of the matter is that as the Steam survey demostrates there's far more GTX580 than there is HD6970. There's very little speculation to be made. Whether you think the HD6970 is much better becomes irrelevant when facts demostrate that gamers do not think like you, as is demostrated by their wallets. Of course the GTX570 and GTX560 ti sells even more, much more.

 

[KashunatoR]

my point exactly. 6970 was aimed at gtx 580 but it ended up competing with the gtx 570. believe me, if amd could have built a stronger single gpu card at that moment, they would have done it, even disregarding the power consum. mind you, they even had the precious advantage or releasing their 6970 later than the gtx 580, but this was as far as they could go. period

 

[Bundy]

* How many serious surveys do you know that were conducted with more than 5% of the targeted population? I'll save your time: NONE.

Agree

OT, I struggle to understand why there would be so much enthusiasm to portray this late delivery issue like it's a football game deserving opposing views. A more mature look at this situation would see that we are all losers when new technology is delayed, no matter what manufacturer.

 

[mauriek]

well..that's why in every statistical survey there's margin of error, and conclusion is only for specific question, whether the survey question is to get statistical data which brand has the most user or which brand has the most sales depend on the question asked in the survey.

 

[Tatty_Two]

I'm not making this into a red/green anything. I'm stating the fact that Nvidia sells more, because that's a fact, it's information. Information to take it as you please. Because it sells more does it meas it's better? No. I'm even trying to imply that? No. But facts are facts. The post I replied to was implying that AMD was selling tons and tons of HD6000 cards more than GTX500, so that they didn't need the new generation as much. I pointed out that they are not selling so many of them, and provided proof. Proof that GTX500 vs HD6000 seems to be leaning in favor of Nvidia, and hence AMD needs to do something to change that trend.

Now regarding the discussion between you and me. You think that Steam survey is not a good source. And you are wrong and I'm telling you why. 50% of games sold digital, most from Steam. By the law of average, and contrary to your assumption, FAR more than 50% of gamers must use Steam, because it only takes 1 game bought through steam to be an Steam user. If this year you bought 100 games and only 1 was from Steam, guess what? You are an Steam user.

And honestly, I still fail to see the relevance of your posts. It doesn't matter if Steam users represent 50%, or 5% of gamers*. To think that a higher proportion of Nvidia users than AMD users get Steam is ridiculous and nonsensical.

* How many serious surveys do you know that were conducted with more than 5% of the targeted population? I'll save your time: NONE.

I am not denying anything you have said, my challenge was simply that you take as fact that a survey which does not include everyone is actually a fact, as I pointed out, it's a good indication but not a fact, I prefer to take facts like the one link you posted earlier, those facts show (along with other recent quarterly reports), that NVidia's market share is in decline and AMD's is increasing, whilst this still places NVidia's with the largest market share of the two, less people currently are buying NVidia cards than they were last year and much of those customers have bought AMD instead..... To be honest I couldnt care who buys what, I have no brand loyality, my last 3 cards have been AMD, my 6 before that were Nvidia....... perhaps I am being a little Anal, and if I am I apologise, however to me facts are facts, what is a fact is that more people who completed the steam survey use NVidia cards, it's a bit like saying 60% of American men drink beer, the other 40% either drink spirits or don't drink at all..... then continuing by saying that the trend therefore must apply to Europe or other parts of the world

 

[Benetanegia]

I am not denying anything you have said, my challenge was simply that you take as fact that a survey which does not include everyone is actually a fact, as I pointed out, it's a good indication but not a fact, I prefer to take facts like the one link you posted earlier, those facts show (along with other recent quarterly reports), that NVidia's market share is in decline and AMD's is increasing, whilst this still places NVidia's with the largest market share of the two, less people currently are buying NVidia cards than they were last year and much of those customers have bought AMD instead..... To be honest I couldnt care who buys what, I have no brand loyality, my last 3 cards have been AMD, my 6 before that were Nvidia....... perhaps I am being a little Anal, and if I am I apologise, however to me facts are facts, what is a fact is that more people who completed the steam survey use NVidia cards, it's a bit like saying 60% of American men drink beer, the other 40% either drink spirits or don't drink at all..... then continuing by saying that the trend therefore must apply to Europe or other parts of the world

The problem with JPR or Mercury Research is that they account for shipments and those include shipments to OEMs, plus they don't make any distinction between high-end and low end, any shipment counts as +1. Shipments do not necessarily equal sales. For example HP may order (random numbers ahead) 2 million AMD cards and 1 million Nvidia cards, but they very well end up selling 600k Nvidia powered PCs and only 400k AMD powered ones. This is just a random and over-exagerated example, but similar things may happen. In that light I cannot take their numbers any more seriously that Steam numbers and I take Steam ones more seriously indeed, because the survey is conducted on final customers (on 30 million active users, 3.5 milion concurrent users) and not the shipments that may later end up on the thrash bin.

And if we judge them by the same standard as you are judging Steam, they are not facts more than Steam survey is anyway. They take into account shipments to the most important retailers and OEMs, but they do not take ALL of them into account either. These analist firms still provide statistics, like anyone else, based on a few retailers and then extrapolate according to expectations. Pretty close to relality, but not 100% sure.

Your last example makes no sense to me either. Arguably americans and europeans (if we even dare to generalise and take europeans as an homogeneous group) have very different tastes and cultures, but Steam does not represent any differentiated demographics compared to gamers in general*, and have members from all countries, ethnics and ages. If there is any distinction, please, tell me which one it is, because there lies the key to our disagreement.

* And ceratinly none regarding GPU election.

 

[KashunatoR]

who cares about market shares? all it matters to me is who delivers THE BEST single gpu card and as little hardware and software problems as possible. it has been nvidia so far with gtx 8800 ultra, gtx 285, gtx 480 and gtx 580. i owned all of this cards. something tells me the tradition will carry on

 

[Tatty_Two]

The problem with JPR or Mercury Research is that they account for shipments and those include shipments to OEMs, plus they don't make any distinction between high-end and low end, any shipment counts as +1. Shipments do not necessarily equal sales. For example HP may order (random numbers ahead) 2 million AMD cards and 1 million Nvidia cards, but they very well end up selling 600k Nvidia powered PCs and only 400k AMD powered ones. This is just a random and over-exagerated example, but similar things may happen. In that light I cannot take their numbers any more seriously that Steam numbers and I take Steam ones more seriously indeed, because the survey is conducted on final customers (on 30 million active users, 3.5 milion concurrent users) and not the shipments that may later end up on the thrash bin.

And if we judge them by the same standard as you are judging Steam, they are not facts more than Steam survey is anyway. They take into account shipments to the most important retailers and OEMs, but they do not take ALL of them into account either. These analist firms still provide statistics, like anyone else, based on a few retailers and then extrapolate according to expectations. Pretty close to relality, but not 100% sure.

Your last example makes no sense to me either. Arguably americans and europeans (if we even dare to generalise and take europeans as an homogeneous group) have very different tastes and cultures, but Steam does not represent any differentiated demographics compared to gamers in general*, and have members from all countries, ethnics and ages. If there is any distinction, please, tell me which one it is, because there lies the key to our disagreement.

* And ceratinly none regarding GPU election.

Lol I quoted it because you provided the link, I agree with you however it's your link not mine

 

[Benetanegia]

Lol I quoted it because you provided the link, I agree with you however it's your link not mine

So the thing is you can't trust statistics?* I understand your point, I just think it's wrong. I mean, W1zzard should just stop doing reviews, because they are never going to be accurate, in fact despite being the reviewer that benches more games, by far, he still cannot bench even a 5% of them, so we should render his reviews useless? His conclusions are "lies" in that they are based on heavily "limited" data? I don't think so, it's wide enough that it represents the mayority of situations out there and hence is valid.

* It must suck to be you, living in this world and distrusting statistics to such a degree. No one takes them 100% seriously but... And what about meds, you must not take meds, they are tested on what 0.001% of human population? And food? The additives may be perjudicial to you! And let's not start talking about cosmetics or bath products. Ugh they must be banned on your home. Just trying to be sarcastic while making a point.

 

[yogurt_21]

while I do agree that according to statistics it doesn't matter how much of the population steam represents, the same trends will likely hold true to all users. Unless of course someone can prove that steam or steam users are biased to one paticular brand. (I seriously doubt it)


but it doesn't matter, I mean why speculate and you can get actual numbers?

http://www.bit-tech.net/news/hardware/2011/05/17/amd-graphics-cards-gain-market-share-from-n/1

The figures, which are based on discrete graphics cards sales in the first quarter of 2011...

while amd is on an upswing, Nvidia simply sells more cards. Stock prices further back that up as investors will flock to the better performer. Right now Nvidia is pulling in nearly double the market cap as AMD.

 

[BazookaJoe]

"Poorer than expected sales"? Do you have any actual evidence for this, or are you just blowing smoke out of your ATI-loving ass?

LOLOLOL!!!

You couldn't be more retarded! Or ignorant.

I HATE my ATI it's one of the worst hardware mistakes I've ever made - and I've said so many times.

 

[Captain.Abrecan]

So the thing is you can't trust statistics?* I understand your point, I just think it's wrong. I mean, W1zzard should just stop doing reviews, because they are never going to be accurate, in fact despite being the reviewer that benches more games, by far, he still cannot bench even a 5% of them, so we should render his reviews useless? His conclusions are "lies" in that they are based on heavily "limited" data? I don't think so, it's wide enough that it represents the mayority of situations out there and hence is valid.

* It must suck to be you, living in this world and distrusting statistics to such a degree. No one takes them 100% seriously but... And what about meds, you must not take meds, they are tested on what 0.001% of human population? And food? The additives may be perjudicial to you! And let's not start talking about cosmetics or bath products. Ugh they must be banned on your home. Just trying to be sarcastic while making a point.

+1 Awesome for making my office laugh out loud

 

[Tatty_Two]

So the thing is you can't trust statistics?* I understand your point, I just think it's wrong. I mean, W1zzard should just stop doing reviews, because they are never going to be accurate, in fact despite being the reviewer that benches more games, by far, he still cannot bench even a 5% of them, so we should render his reviews useless? His conclusions are "lies" in that they are based on heavily "limited" data? I don't think so, it's wide enough that it represents the mayority of situations out there and hence is valid.

* It must suck to be you, living in this world and distrusting statistics to such a degree. No one takes them 100% seriously but... And what about meds, you must not take meds, they are tested on what 0.001% of human population? And food? The additives may be perjudicial to you! And let's not start talking about cosmetics or bath products. Ugh they must be banned on your home. Just trying to be sarcastic while making a point.

I don't distrust statistics at all, and cmon, what has meds got to do with it, all I distrust is anyone that states the word "fact" without facts to back it up, and yes of course I buy things off the back of a review, difference is there are reviewers we trust and reviews we don't, same applies to me, you and everyone else..... but thats irrelivent, reviews are not facts, their reviews! Ultimatly we agree to disagree..... lets move on as we will just continue to go around in circles.

 

[Benetanegia]

I don't distrust statistics at all, and cmon, what has meds got to do with it, all I distrust is anyone that states the word "fact" without facts to back it up, and yes of course I buy things off the back of a review, difference is there are reviewers we trust and reviews we don't, same applies to me, you and everyone else..... but thats irrelivent, reviews are not facts, their reviews! Ultimatly we agree to disagree..... lets move on as we will just continue to go around in circles.

So the conclusion to all this is that I wasted my precious time to semantics? LOOOOOOOOLOLOLOL.