Another AMD leaked slide shows upcoming BARTS is based on Radeon 5xxx architecture...

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Idontcare

Elite Member
Oct 10, 1999
21,118
58
91
The archnemesis of Wreckage! :biggrin: I can't wait to see both of them duke it out.

lol, +1 to that.

ANd I'll make the popcorn. :)

Silver is way more moderate than Wreckage. No idea how you can even compare those two.

Yeah, but he's just getting started. Give him some time to moult.

Gentlemen,

No posting your opinion of fellow members if that opinion is negative, dismissive, derogatory or disrespectful. :thumbsdown: (and that includes sarcastically positive comments as well)

Making genuinely positive statements regarding your fellow forum members is acceptable. :thumbsup:

Please be mindful of these expectations in the technical forums.

All the quoted posts listed above would ordinarily garner an infraction for personal attacks/insults/call-outs.

Today you get a friendly warning, find something nice to say or don't bother posting it to begin with. Next time it will be infractions.

Regards,
Moderator Idontcare
 

Seero

Golden Member
Nov 4, 2009
1,456
0
0
There is no indication of a declining rate of sales (at least not in the steam data and not for the last few months) for the 5xxx series in the DX11 segment. There is a decline of market share which is not the same.

Since more DX11 cards started to appear the 5800 series has been on a decline of DX11 market share and the 5700 series went on a decline shortly after their launch too.
I have mis-leaded you when I use 5xxx. What I really mean is 58xx/57xx. They are shrinking in terms of sales. I am not saying that are not selling, I am saying that they are not selling as fast as before april.

The chart starts at april, and assuming no many people get there hands on the chip at September, the average percentage change for 5800 is 2.34%/7=.334%. For 5700, it should be even better since it was out at october.

This is logical as we knew at the time there are almost no competitions except some IGPs from Intel. After the initial the 57xx and 58xx have competition within the Dx11 segment. Some are ATI's mobile chipsets, IGPs, and 56xx/55xx. It really doesn't take a rocket scientist to figure out that 58xx/57xx isn't selling as well as before April, and based on the grade 4 math above, the numbers do agree with that.

Combining 57xx/58xx they actually took 4.77% of the market in 6 months, that is a lot of sells. Do not forget they never really have a price adjustment, meaning that they are really milking from them. In 11 months time, they milked 6.53% out of the market in total with the 2 series. That is a lot of sales. However, the rate of sales isn't as strong as before.

The average rate of growth on the chart for 58xx is .207%, down from .344%, indicating a drop of sales. Now that isn't an indication of AMD losing share from the market though.

Another indication is the rate of growth of Dx11 video card is more or less constant, yet the rate of growth of 58xx is decreasing. The former number indicate that the need of Dx11 cards have not been saturated, yet the need of 58xx isn't as strong as before.

None of these indicate any relation between the sales of 460 and the sales of 58xx/57xx. Independently speak however, 460 appears to be doing very well, and this forum won't be surprised. 58xx/57xx is slowing down, which this forum is also not surprised. In fact, we expected it.

What is going to hurt AMD isn't 460, but 6xxx. I will not expect 6xxx will come as strong as 5xxx as they came in where their are no competition. However, what exactly will happen if 6xxx doesn't live up to expectation? We know it will sell well if it is twice as fast, and we know it won't sell well if it is a brick. However, the result won't be so extreme. What if the first 6xxx performs like the 5850, will it sell well? or will it simply take over the portion which belongs to 5850?

AMD don't die even if 6xxx are bricks. They can simply adjust the price of 5xxx and they will still be good for a while. The difference is, doing better than before, or not as good.
 
Last edited:

Outrage

Senior member
Oct 9, 1999
217
1
0
The first paragraph of http://www.anandtech.com/show/2549

G80 = the first generation of the GT architecture.
G92 = the second generation of the GT architecture.
Each of the has a family and also belongs to the GT achitecture family.

G80 = the first generation of the GT architecture.
G92 = the first generation of the GT architecture.
GT200 = the second generation of the GT architecture.

Thats from the first paragraph. The next fermi generation will be called GF2XX
 

GaiaHunter

Diamond Member
Jul 13, 2008
3,628
158
106
I have mis-leaded you when I use 5xxx. What I really mean is 58xx/57xx. They are shrinking in terms of sales. I am not saying that are not selling, I am saying that they are not selling as fast as before april.

The chart starts at april, and assuming no many people get there hands on the chip at September, the average percentage change for 5800 is 2.34%/7=.334%. For 5700, it should be even better since it was out at october.

This is logical as we knew at the time there are almost no competitions except some IGPs from Intel. After the initial the 57xx and 58xx have competition within the Dx11 segment. Some are ATI's mobile chipsets, IGPs, and 56xx/55xx. It really doesn't take a rocket scientist to figure out that 58xx/57xx isn't selling as well as before April, and based on the grade 4 math above, the numbers do agree with that.

Combining 57xx/58xx they actually took 4.77% of the market in 6 months, that is a lot of sells. Do not forget they never really have a price adjustment, meaning that they are really milking from them. In 11 months time, they milked 6.53% out of the market in total with the 2 series. That is a lot of sales. However, the rate of sales isn't as strong as before.

The average rate of growth on the chart for 58xx is .207%, down from .344%, indicating a drop of sales. Now that isn't an indication of AMD losing share from the market though.

Another indication is the rate of growth of Dx11 video card is more or less constant, yet the rate of growth of 58xx is decreasing. The former number indicate that the need of Dx11 cards have not been saturated, yet the need of 58xx isn't as strong as before.

None of these indicate any relation between the sales of 460 and the sales of 58xx/57xx. Independently speak however, 460 appears to be doing very well, and this forum won't be surprised. 58xx/57xx is slowing down, which this forum is also not surprised. In fact, we expected it.

What is going to hurt AMD isn't 460, but 6xxx. I will not expect 6xxx will come as strong as 5xxx as they came in where their are no competition. However, what exactly will happen if 6xxx doesn't live up to expectation? We know it will sell well if it is twice as fast, and we know it won't sell well if it is a brick. However, the result won't be so extreme. What if the first 6xxx performs like the 5850, will it sell well? or will it simply take over the portion which belongs to 5850?

AMD don't die even if 6xxx are bricks. They can simply adjust the price of 5xxx and they will still be good for a while. The difference is, doing better than before, or not as good.

100 units sold 1st month.
100 units sold 2nd month. 100% increase.
100 units sold 3rd month. 50% increase.
100 units sold 4rd month. 25% increase.

There is a situation where the monthly sales number is flat, but the growth % of sales is decelerating.

If the number of steam accounts is 25 million, the steam DX11 market is 2542500 units, from which 791988 units are series 5800 and 840296 units are 5700 series. 87208 are GTX460. That is 10:1 for each individual 5800/5700 series or 20:1 when combined.

If the GTX460 keeps selling ~80k cards a month it is going to take 10 months to reach 5800 and 5700 levels.

So lets not forget that the 5800 and 5700 series are close to 1 year old and the 5800 are still selling at around launch prices. Any people interested in 5800 and 5700 series has long bought one.

This doesn't mean 5800 and 5700 series sales is decelerating or that GTX460 isn't selling good. On the other hand, unless you have hard sale numbers, you can't prove either 5800 and 5700 are selling less and that is because GTX460 or whatever card (of course if NVIDIA had no DX11 cards, all the DX11 cards would be AMD).
 

Seero

Golden Member
Nov 4, 2009
1,456
0
0
100 units sold 1st month.
100 units sold 2nd month. 100% increase.
100 units sold 3rd month. 50% increase.
100 units sold 4rd month. 25% increase.

There is a situation where the monthly sales number is flat, but the growth % of sales is decelerating.
I disagree. Lets say i have 10 green beans in the bucket today, and I replace 1 green bean with a red bean a day. The rate of replacement is 10% daily. This is like the overall video card where people replace their old card for a new one.
Lets look at the case where I start with 1 green bean in the bucket and add a red bean everyday for 3 days. The proportion of green/red starts at 50%, then 66%, then 75% in 3 days. The growth of proportion of red is 50%, 16%, and 9%. This is what the Dx11 proportion should look like because very few people is replacing their Dx11 card with another Dx11 card.
In the all video card segment, the rate of growth of the 58xx/57xx dropped. In the Dx 11 segment, it rate of growth he 58xx/57xx is negative. Both of them indicated that it is a shrink of growth.
If the number of steam accounts is 25 million, the steam DX11 market is 2542500 units, from which 791988 units are series 5800 and 840296 units are 5700 series. 87208 are GTX460. That is 10:1 for each individual 5800/5700 series or 20:1 when combined.

If the GTX460 keeps selling ~80k cards a month it is going to take 10 months to reach 5800 and 5700 levels.
If the sales of 58xx/57xx stops and the growth of 460 stops, then yes 460 will reach 5800 and 5700 levels in 10 months. However, this is not a practical scenario.

There is only one flicking month of data from 460 so I can say whatever theory I want but that isn't the point. However, with the price drop it is likely that the growth of sales will increase.

So lets not forget that the 5800 and 5700 series are close to 1 year old and the 5800 are still selling at around launch prices. Any people interested in 5800 and 5700 series has long bought one.
I agree, that those series has made enough money even if AMD sells the reminding stock at 1 dollar (assuming that stop producing more).

... On the other hand, unless you have hard sale numbers, you can't prove either 5800 and 5700 are selling less and that is because GTX460 or whatever card (of course if NVIDIA had no DX11 cards, all the DX11 cards would be AMD).
I don't think I can prove even if i have the sales numbers. Maybe some Actuary can.
 

beginner99

Diamond Member
Jun 2, 2009
5,208
1,580
136
they are forced to release an "upgraded" version of Radeon 5xxxx series just to compete with the GTX 460 which is giving AMD tons of headaches...

lol.

The 460 is a little slower than a 5850 and a little cheaper (but more power hungry). All in all it might be the better buy but the difference is marginal at best. 5850 prices are probably still high because of ongoing demand and not because AMD ain't lowering them.
Anyway a 5850 might actually be cheaper to produce than a 460. Who knows?
 

Schmide

Diamond Member
Mar 7, 2002
5,581
712
126
I disagree. Lets say i have 10 green beans in the bucket today, and I replace 1 green bean with a red bean a day. The rate of replacement is 10% daily. This is like the overall video card where people replace their old card for a new one.
Lets look at the case where I start with 1 green bean in the bucket and add a red bean everyday for 3 days. The proportion of green/red starts at 50%, then 66%, then 75% in 3 days. The growth of proportion of red is 50%, 16%, and 9%. This is what the Dx11 proportion should look like because very few people is replacing their Dx11 card with another Dx11 card.
In the all video card segment, the rate of growth of the 58xx/57xx dropped. In the Dx 11 segment, it rate of growth he 58xx/57xx is negative. Both of them indicated that it is a shrink of growth.

You do understand, and this has been happening to nVidia with its g92 segment, if you have the lions share of a segment and the whole segment grows by an average amount, you will see a relative negative number because the total pie has increased more than your relative growth even when you too have increased your total number.

I'm not discounting the growth of the 460 in any way, 0-3.5 is a tremendous number, but the negative numbers in the 5700/5800 are just because of increasing domains.
 

Lonyo

Lifer
Aug 10, 2002
21,939
6
81
Please stop using and caring about Steam numbers, they are meaningless when it comes to sales.

The best they can do is suggest what level of hardware a reasonable selection of gamers have.
They can't be used to show sales of graphics cards.
 

PC_Guy

Banned
Sep 16, 2010
2
0
0
http://www.fudzilla.com/graphics/graphics/graphics/amd-barts-specs-leaked

"Barts XT will feature 12 SIMDs, 960 shaders, 48 texture units and 32 ROPs. For comparison, Juniper XT packs 800 shaders and 40 texture units, so Barts will have 20 percent more of everything. The core should end up clocked at 850MHz and the TDP is rated at over 150W, which is quite a bit more than the HD 5770.

Barts PRO packs 800 shaders, 40 texture units and 32 ROPs, much like the current Juniper XT, but it will also have a wider 256-bus. It will be clocked between 700MHz and 725MHz and its TDP should end up below 150W.

AMD is pitting both boards against Nvidia’s GTX 460: Barts PRO will take on the 768MB version of Nvidia’s GTX 460, while Barts XT should fight the 1GB version. Judging by the spec, Barts boards should be able to outperform the GTX 460 and we wouldn't be surprised if they match HD 5830 or even 5850 performance in some scenarios, thanks to higher clocks and the 256-bit bus."
 

Seero

Golden Member
Nov 4, 2009
1,456
0
0
Please stop using and caring about Steam numbers, they are meaningless when it comes to sales.

The best they can do is suggest what level of hardware a reasonable selection of gamers have.
They can't be used to show sales of graphics cards.
Maybe you want to read the posts in the thread next time before posting.

Sampling from wikipedia.org

Unless you can prove(unless an indication) that it is bias to whatever we are discussing, your post is meaningless.
 

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
Please stop using and caring about Steam numbers, they are meaningless when it comes to sales.

The best they can do is suggest what level of hardware a reasonable selection of gamers have.
They can't be used to show sales of graphics cards.

My thoughts exactly. Somehow people are taking Steam results as indicative of overall global market sales. :rolleyes: We don't know anything about the sale prices or sales of these cards in Europe, Asia, Oceania, South America, etc.
 

Daedalus685

Golden Member
Nov 12, 2009
1,386
1
0
Maybe you want to read the posts in the thread next time before posting.

Sampling from wikipedia.org

Unless you can prove(unless an indication) that it is bias to whatever we are discussing, your post is meaningless.
It can't be correlated to sales figures because it is a changing sample of users. Thus there is no separation between those that bought a 5870 in January but installed steam today and those who already had steam (and appeared on the survey) but just bought a 5870.

It is only a rough idea on current market appearances and any derivation of sales fugues is not valid from any sample unless you maintain the exact same sample over time (steam does not). The best you can say is how the given proportion of cards might be changing with time but you can't correlate that with new sales to any accuracy as the sample is not static. So while it is a good representation of what a population uses as a card, there is no way to pull from that sales in a given time period as any increase in card units can be equivalently explained by an increase in sample size.
 

tviceman

Diamond Member
Mar 25, 2008
6,734
514
126
www.facebook.com
If this is an official slide, to me it's an indication that these parts won't exceed the gtx460 768/1gb in performance. The touting mantra is "better performance per watt" as opposed to "better performance with less power draw" or "equal performance with less power draw". If this ends up being the case, very few people will actually care that it's 95% the performance of a gtx460 while having 20% less TDP. Bottom line, if the new 6750/6770 can't noticeably outperform a gtx460 768/1gb in most cases, and considering the current pricing of the gtx460's, to me this part is destined to be a lackluster release.
 

GaiaHunter

Diamond Member
Jul 13, 2008
3,628
158
106
If this is an official slide, to me it's an indication that these parts won't exceed the gtx460 768/1gb in performance. The touting mantra is "better performance per watt" as opposed to "better performance with less power draw" or "equal performance with less power draw". If this ends up being the case, very few people will actually care that it's 95% the performance of a gtx460 while having 20% less TDP. Bottom line, if the new 6750/6770 can't noticeably outperform a gtx460 768/1gb in most cases, and considering the current pricing of the gtx460's, to me this part is destined to be a lackluster release.

It is that official slide that is from power point in EDIT mode.
 

Skurge

Diamond Member
Aug 17, 2009
5,195
1
71
It is that official slide that is from power point in EDIT mode.

Yes, I don't ever remember nV or AMD using the term "Shader" in their slides. They use stuff like Cuda cores or stream processors or stream cores. That slide just looks bogus or made up to me.
 

railven

Diamond Member
Mar 25, 2010
6,604
561
126
It can't be correlated to sales figures because it is a changing sample of users. Thus there is no separation between those that bought a 5870 in January but installed steam today and those who already had steam (and appeared on the survey) but just bought a 5870.

It is only a rough idea on current market appearances and any derivation of sales fugues is not valid from any sample unless you maintain the exact same sample over time (steam does not). The best you can say is how the given proportion of cards might be changing with time but you can't correlate that with new sales to any accuracy as the sample is not static. So while it is a good representation of what a population uses as a card, there is no way to pull from that sales in a given time period as any increase in card units can be equivalently explained by an increase in sample size.

To add to this, I have five steam accounts (don't ask why, harvested them from friends :D) and I'm sure each one of my accounts shows a HD 5870 in the survey for the month of April (when I bought it) or May (when I installed Steam on a fresh drive) but trust me as much as I'd like - I don't have 5 HD 5870s haha.

Those are variables I'm sure people have to take into consideration. Also, does the survey affect migration? I've logged into my steam accounts on my girlfriend's PC - would it register her GTX 460 for my user info in the surveys? If so, that is another 4 phantom GTX 460's.
 

Lonyo

Lifer
Aug 10, 2002
21,939
6
81
Maybe you want to read the posts in the thread next time before posting.

Sampling from wikipedia.org

Unless you can prove(unless an indication) that it is bias to whatever we are discussing, your post is meaningless.

How about the fact that some cards come bundled with games that require Steam?
That would put them as being likely over-represented in the survey.
If you get a free game with your graphics card, you are more likely to install Steam, favouring that card over others.
Steam isn't a random sample, it's not a representative sample, it's a biased sample in many ways and for many reasons, the above is just one example.
It's also biased towards higher end cards on the (assumed) basis that people with "gaming" cards are more likely to... game. If they are more likely to game, they are more likely to have Steam. This makes Steam biased towards gaming type cards, and thus not reflecting the overall market. It's never been the case (based on the numbers ATI have released) that the 5700 and 5800 series have had equal sales, yet they have similar Steam marketshares. That means it's biased against the low end and not representative.

So, two specific examples of bias as well as a general comment about it being unrepresentative.
Plus all the other things like self reporting issues and limited sample size.

You specifically talk about "negative growth" of the HD5700 cards, which are one of those which is misreported in terms of general sales, based on matching Steam marketshare % with the hard sales numbers reported by ATI back in January or so (when they were the only DX11 cards out) and Steam showing something like 55/45 split between the 5800/5700 while hard numbers from ATI showed it being 38/62 5800/5700, or similar. It's in posts I've made previously in other threads long ago while saying Steam is useless for these sorts of discussions.
 

tviceman

Diamond Member
Mar 25, 2008
6,734
514
126
www.facebook.com
It is that official slide that is from power point in EDIT mode.

Yes, I don't ever remember nV or AMD using the term "Shader" in their slides. They use stuff like Cuda cores or stream processors or stream cores. That slide just looks bogus or made up to me.

Yeah I'm not sure what is real information and what is made up stuff, which is why I qualified my statements with the word "if" many times. However I think it's interesting that most people are quicker and more apt to believe the rumors/leaks that paint these upcoming products in a better light rather than rumors/leaks that tend to weigh down anticipation.

I've said several times in regards to the benchmark leaks and slide leaks that IF the 6870 is actually 25% faster than a gtx480 in DX11 tessellation and consumes less power it'll be a hell of a product, just like I'm saying now that if 6750/6770 parts can't convincingly beat the gtx460's they'll be disappointing. I know it's all still just speculation, and speculation can be fun, but nevertheless interesting how different rumors, even if one is just as believable as the other, are perceived and accepted.
 

Skurge

Diamond Member
Aug 17, 2009
5,195
1
71
Yeah I'm not sure what is real information and what is made up stuff, which is why I qualified my statements with the word "if" many times. However I think it's interesting that most people are quicker and more apt to believe the rumors/leaks that paint these upcoming products in a better light rather than rumors/leaks that tend to weigh down anticipation.

I've said several times in regards to the benchmark leaks and slide leaks that IF the 6870 is actually 25% faster than a gtx480 in DX11 tessellation and consumes less power it'll be a hell of a product, just like I'm saying now that if 6750/6770 parts can't convincingly beat the gtx460's they'll be disappointing. I know it's all still just speculation, and speculation can be fun, but nevertheless interesting how different rumors, even if one is just as believable as the other, are perceived and accepted.

True, every word.

I want the cards to be good, I want them to make all the current cards obsolete, I want progress. So you can forgive for wanting to believe the more positive rumors than the negative ones.
 

GaiaHunter

Diamond Member
Jul 13, 2008
3,628
158
106
Yeah I'm not sure what is real information and what is made up stuff, which is why I qualified my statements with the word "if" many times. However I think it's interesting that most people are quicker and more apt to believe the rumors/leaks that paint these upcoming products in a better light rather than rumors/leaks that tend to weigh down anticipation.

I've said several times in regards to the benchmark leaks and slide leaks that IF the 6870 is actually 25% faster than a gtx480 in DX11 tessellation and consumes less power it'll be a hell of a product, just like I'm saying now that if 6750/6770 parts can't convincingly beat the gtx460's they'll be disappointing. I know it's all still just speculation, and speculation can be fun, but nevertheless interesting how different rumors, even if one is just as believable as the other, are perceived and accepted.

Did I say anything incorrect in my post?

How can it be an official slide if it is in edit mode?
 

Seero

Golden Member
Nov 4, 2009
1,456
0
0
How about the fact that some cards come bundled with games that require Steam?
That would put them as being likely over-represented in the survey.
If you get a free game with your graphics card, you are more likely to install Steam, favouring that card over others.
What exactly are you trying to say? Are you trying to say that more Nvidia user use steam?
Steam isn't a random sample, it's not a representative sample, it's a biased sample in many ways and for many reasons, the above is just one example.
Please explain what you are trying to say first before you use it as an example.
It's also biased towards higher end cards on the (assumed) basis that people with "gaming" cards are more likely to... game.
If you are trying to compare 2 cards, one for gaming, and one isn't, then you may have a point. My analysis only consist of looking a card/series individually.
If they are more likely to game, they are more likely to have Steam. This makes Steam biased towards gaming type cards, and thus not reflecting the overall market. It's never been the case (based on the numbers ATI have released) that the 5700 and 5800 series have had equal sales, yet they have similar Steam marketshares. That means it's biased against the low end and not representative.
I can agree that steam may be biased towards gamers, but that does not affect the analysis itself because:
a) The target market of these cards are for gaming.
b) 5800/5700 as well as 460 are all good for gaming.
c) ATI have just added auto driver update with steam, meaning that the scope of the group may increase.

So, two specific examples of bias as well as a general comment about it being unrepresentative.
Plus all the other things like self reporting issues and limited sample size.

You specifically talk about "negative growth" of the HD5700 cards, which are one of those which is misreported in terms of general sales, based on matching Steam marketshare % with the hard sales numbers reported by ATI back in January or so (when they were the only DX11 cards out) and Steam showing something like 55/45 split between the 5800/5700 while hard numbers from ATI showed it being 38/62 5800/5700, or similar. It's in posts I've made previously in other threads long ago while saying Steam is useless for these sorts of discussions.
Again, I didn't compare any 2 cards/series, but simply looking at the relative movement of share to the "entire video card segment" and the "Dx11 card segment."
For example, the percentage of share doesn't reflect the actual share. That means it does not mean ATI share > Nvidia share just because the total percentage of ATI cards > the total percentage of Nvidia cards on that survey. However, based on the monthly difference of a specific card/series, I do have a good indication of whether a series is doing well or not. You may argue that it may not be completely accurate, but that happens to any statistics and survey.

It is ignorant to believe that there are no data which we can find out the trend of sales. There are job fields just for these matters like marketing, analysist and actuary. How to people claim "smoking causes cancer?" Is it a wild guess, rumors, or a conclusion made based upon the statistics? Has any test done based on a randomly selected group all over the world on every age, sex, height, weight, and blood type, and put half of them in a smoke area for 30 years? No, they use mouse instead and suffocate the smoking group with smoke several times an hour and ask cancer patient and see if they smoke or not and for how long if they do. If you strongly believe that those are all completely BS and none of them knows what they are talking about, then maybe you are just being stubborn.

Again, it is okay to challenge the conclusion or the method which is used to derive it. I am not saying you should listen and simply trust BS, I am saying that you guys can do better than just say "It is dump to try."
 

GaiaHunter

Diamond Member
Jul 13, 2008
3,628
158
106
What exactly are you trying to say? Are you trying to say that more Nvidia user use steam?

AMD also included Dirt2 (if I'm not mistaken) and some AvP(?) with the initial release of the 5800/5700(?) series.

I think it is a valid point towards the inaccuracies of steam to predict monthly sales.

Although steam statistics have some value, for example, it is easy to see GTX460 is a much more popular card thant the GTX480/470/465 and probably as popular as 5800/5700 series.
 

Schmide

Diamond Member
Mar 7, 2002
5,581
712
126
Again, I didn't compare any 2 cards/series, but simply looking at the relative movement of share to the "entire video card segment" and the "Dx11 card segment."
For example, the percentage of share doesn't reflect the actual share. That means it does not mean ATI share > Nvidia share just because the total percentage of ATI cards > the total percentage of Nvidia cards on that survey. However, based on the monthly difference of a specific card/series, I do have a good indication of whether a series is doing well or not. You may argue that it may not be completely accurate, but that happens to any statistics and survey.

I do hate to say it, but you're making a mountain out of a mole hill. No Fermi card has yet to reach the 0.5% share to make it on the main board. Rough analysis says they're right on the cusp and the 460 will be the first to make it some time in this month. Total Fremi adoption is probably still in the 2-3% range.

It is very hard to gauge long term sales on early changes in percentages. If you think you can, i have some penny stocks to sell ya.

I have no doubt the 460 will be the best selling card of nVidia's current lineup, but long term trends cannot be guessed by early adoption rates.
 

Seero

Golden Member
Nov 4, 2009
1,456
0
0
I do hate to say it, but you're making a mountain out of a mole hill. No Fermi card has yet to reach the 0.5% share to make it on the main board. Rough analysis says they're right on the cusp and the 460 will be the first to make it some time in this month. Total Fremi adoption is probably still in the 2-3% range.

It is very hard to gauge long term sales on early changes in percentages. If you think you can, i have some penny stocks to sell ya.

I have no doubt the 460 will be the best selling card of nVidia's current lineup, but long term trends cannot be guessed by early adoption rates.
What mountain are you speaking of? What long term sales? What do you think I am thinking?

All I have ever predicted in this thread is, "the sales of 460 will likely be better due to the price drop." If that sounds like a myth to you, then I have nothing to say.
...
There is only one flicking month of data from 460 so I can say whatever theory I want but that isn't the point. However, with the price drop it is likely that the growth of sales will increase...
 
Last edited:

Schmide

Diamond Member
Mar 7, 2002
5,581
712
126
What mountain are you speaking of? What long term sales? What do you think I am thinking?

It's a figure of speech? Meaning you're looking at a small trend and trying to show something bigger.

...

I have no idea what you're thinking most of the time.

It is ignorant to believe that there are no data which we can find out the trend of sales.

It is even more ignorant to look at one blip and claim trend. It is more accurate to say the 460 is priced very well for its performance and should take its relative share of the segment.