Wow Recent survey results.

[Golgatha]

I'd lay money the 5200 series doing so well is the same phenomenon I've seen three times at Best Buy and those ridiculously pointless 256MB 6200 OC BFG cards.

What a great value! I get 256MB of RAM, it's overclocked, and cheaper than that overpriced 128MB 6600GT. Woohoo!!! All the sheep want are big numbers, words like "overclocked" on the box, and a cheap price. The vast majority of people are too lazy to actually research a major purchase IMO, which is why the graphics card companies release such pointless doody.

 

[Matthias99]

Originally posted by: BenSkywalker

To console gaming, yes. In comparison to the 'niche' of 'enthusiast PC gamers', IMO no.

WoW dwarfs it also- and that is never scheduled to hit consoles. Even when limiting it to FPSs, CS is dwarfed by the original HL. Online FPSs is a very small niche even when looking at PC gaming.

Um, HL/CS players are included in this survey. WoW has in the area of 5-6 million subscribers, and even the numbers you gave showed just HL selling, what, 10 million copies? That doesn't seem like it is 'dwarfed' to me. I don't feel like I am going to get far trying to convince you.

In all honesty- how do you expect anyone to make any sense out of your appraisal of the situation when you include the x700- which came out after the 6600 series and cost as much as the GT varriant? Because the GT was so far superior to what ATi was offering at that price point? Is that the logic behind it? I'm going to ignore the hands down most popular part from this survey because it doesn't agree with what I want to say. Statistical analysis can show you how to twist the numbers to say what you want them to, but even the most convuluted method of twisting wouldn't allow what you are doing.

Gee, Ben, tell me how you really feel. Could you try posting with a little more melodrama?

It's very, very hard from these numbers to break things down nicely like you want to, because there is nothing here that indicates when the cards were bought and what price was paid for them. And it seems you disagree with my interpretation. Let me try to explain:

IMO, the X600/700 basically competed (well) with the FX5500/5600/5700 and (poorly) with the GF6600, and the X800/X800GTO competed with the 6600GT/6800NU (although there was a gap in time where ATI basically had no direct competitor in that price segment). But the X800 numbers are lumped in with the X800XL/XT, and the 6800NUs are in with the 6800GT/Ultras.

For reference, here are the numbers I'm looking at:

NVIDIA GeForce4 Series 21,008
NVIDIA GeForce FX 5500 15,973
NVIDIA GeForce FX 5600 11,806
NVIDIA GeForce FX 5700 Series 17,319
NVIDIA GeForce 6200 Series 13,630
NVIDIA GeForce 6600 Series 78,439
NVIDIA GeForce 6800 Series 47,413

ATI Radeon 9500/9700 9,489
ATI Radeon 9550 13,650
ATI Radeon 9600 Series 56,484
ATI Radeon 9800 Series 43,801
ATI Radeon X600 11,889
ATI Radeon X700 Series 16,215
ATI Radeon X800 Series 40,302

That covers pretty much the entirety of the mid- to high-end of the previous hardware generation. If you total them up, you get:

NVIDIA: ~205,000
ATI: ~191,000

But what I don't know is the breakdown of the "X800" and "6800" buckets -- the 6800NU and X800[GTO] could maybe reasonably be considered "midrange" cards (we're not using the same definitions, apparently), but the X800XL/XT and 6800GT/Ultra are not. To me, it seemed like there was a definite step up in both price and performance between the 6600/X700 and the 6600GT/6800NU/X800 -- enough that they should not be considered together. Maybe that wasn't the best way to analyze it, but you don't have to rip me and accuse me of trying to twist the numbers to get the result I wanted.

It seems pretty evenly split to me no matter how you slice it; I'm not seeing the midrange NVIDIA dominance that was being claimed earlier. ATI got behind quite markedly at the higher end after the launch of the 7800GT, but the X800/X800GTO has been a pretty solid midrange alternative to the 6600GT since its launch.

 

[gramboh]

We know HL/HL2 is well over 10M copies sold because there are over 12M unique Steam accounts in existence (steamID is sequential, you will see users on servers with XX,XXX,XXX as their ID).

There are always over 200,000 people playing Steam games at any given point in time, sometimes over 300,000.

How are these not popular PC games?

I don't know much about WoW, isn't it ~5k players per server and about 80 servers? So that would be 400,000 people, more than Steam yes, but still in the same ballpark. CS/HL1 games are the most popular FPS by far.

 

[Matthias99]

Originally posted by: gramboh
How are these not popular PC games?

Because Ben says so, duh!

I don't know much about WoW, isn't it ~5k players per server and about 80 servers? So that would be 400,000 people, more than Steam yes, but still in the same ballpark. CS/HL1 games are the most popular FPS by far.

I think there are way more than 5K accounts per server. Post-Xmas, I was on one of the busier servers, and there were 1,000+ people in the logon queue during peak hours on just that one server.

Blizzard claims to have 5-6 million 'subscribers', but I dont know if that is currently paying subscribers or just copies sold and activated (but not necessarily still playing).

 

[fliguy84]

According to Xfire, the total hours people play CS and CS:S is more than the total hour of people playing Wow. So that must mean CS and CS:S (Steam) is the most popular online game. So I think the steam survey is quite valid

http://www.xfire.com/

 

[imported_Crusader]

OH
and just for the record.. Valve survey results have proven very reliable in the past to correlate well with actual overall sales.
Conclusions on the overall market CAN and HAVE and SHOULD be drawn from this survey.. its the best indicator (and most accurate historically) we have beyond ZipZoomfly/Monarch/Newegg from telling us their sales statistics.

Sorry that many of you (ATI fanboys mainly) do not like this but that is just the cold hard reality of the situation!

 

[Topweasel]

WE all know That Nvidia went through a rough patch with the FX series cards, thats not the point. The Point I and I think most people were trying to show was the Down turn that ATI has been going through since (and this was the surprise to me) the X800 (and there fore X300, X600 and X700) series. I ad always thought that everything was basically even and the X1800 wast a blip. But in reality (or as much of it that this survey can convey) ATI was only doing OK after it tremendously successfull 9000 series, It lost quite abit of ground during the X800 days compared to the 9000 days, and the X1800 screwup appears to have cost them dearly.

 

[BenSkywalker]

My point was more to the fact that HL and HL2 are popular enough games amoung players of any type of genre and is usually the most graphically intensive genre to say that this is a fair, maybe a little low end biased, but over all fair look at the current status of the video and CPU usage of people who play computer games in general.

When I have said anything remotely coming close to being in the league of arguing this point? When I have I questioned the validity of this survey? I have taken issue with what people are trying to draw from it- that is all.

Um, HL/CS players are included in this survey. WoW has in the area of 5-6 million subscribers, and even the numbers you gave showed just HL selling, what, 10 million copies? That doesn't seem like it is 'dwarfed' to me. I don't feel like I am going to get far trying to convince you.

And how does that in any way relate to current CS players which was very explicitly what we were talking about? You explain it to me, because I'm not seeing it.

It's very, very hard from these numbers to break things down nicely like you want to

It isn't hard for anyone who pays even a passing bit of attention to the vid card market. The x700 was explicitly released as the direct competitor to the 6600GT- it was in price point and it was released after it. It is very cut and dry.

IMO, the X600/700 basically competed (well) with the FX5500/5600/5700 and (poorly) with the GF6600, and the X800/X800GTO competed with the 6600GT/6800NU (although there was a gap in time where ATI basically had no direct competitor in that price segment). But the X800 numbers are lumped in with the X800XL/XT, and the 6800NUs are in with the 6800GT/Ultras.

The x800/x800GTO also both had direct competitors- the 6800 and 6800GS respectively, that were very clearly defined and not a hint of the companies trying to hide that fact. In fact, they were quite clear on exactly where they were positioned.

To me, it seemed like there was a definite step up in both price and performance between the 6600/X700 and the 6600GT/6800NU/X800

The x700Pro was the same price as the 6600GT, they are still within $20 after it has been painfully obvious that the part was never competitive for some time.

I'm not seeing the midrange NVIDIA dominance that was being claimed earlier.

In simple market terms, for anything close to current gen it comes down to this-

NVIDIA GeForce 6600 Series 78,439

ATI Radeon X600 11,889
ATI Radeon X700 Series 16,215

That is the straighforward both IHV endorsed midrange segment. Even if you include all of the x800 parts(we'll call it 41K) the 6600 series still comes out ahead. That is the midrange dominance people were talking about earlier. It isn't remotely close.

According to Xfire, the total hours people play CS and CS:S is more than the total hour of people playing Wow.

That only covers those using XFire- not something your typical MMO player would be interested in(nor your typical gamer in general). If the claims that there are 200K CS players online 24 hours a day that would be 288,000,000 minutes- according to XFire it is about 1% of that.

 

[Topweasel]

Originally posted by: BenSkywalker

*Snip*

According to Xfire, the total hours people play CS and CS:S is more than the total hour of people playing Wow.

That only covers those using XFire- not something your typical MMO player would be interested in(nor your typical gamer in general). If the claims that there are 200K CS players online 24 hours a day that would be 288,000,000 minutes- according to XFire it is about 1% of that.

Quite worng on the Xfire thing, I have several friends who only play RTSs ad MMOs and they run Xfire all the time. I am about the only consistent FPS player in that little group and I am not a big fan of using Xfire, to me its jus away for them to tell I am on and calling me a wuss for not playing the MMO they are playing at the time or allow them to go on and on about the size of the group they are in. So just like the Survey, Xfire becomes a program limited in its usage in the gaming community on a whole, but be large enough sampling of the community as a whole to give a realistic Idea of people gaming activities. As for the one percent thing, keep in mind most polling and statistics comes from pools that are much lower then 1 percent. TV ratings are based on thousands of viewers in markets of several million, with Digital boxes its becoming a little different, but then again with Tivo and DVRs those numbers are influx because their is a real question about whether they are watching live, if they would watch live if that was the only option, and even the ones who would whether or not they forward through commercials.

Honestly these two Surveys I think are better representation of what we are looking for then some of surveys used as truths for the american public. Some seem like they are polling 10 thousand peole in Birmingham, Mi or Beverly Hills, CA to base average income. Or the Opposite like going to Detroit or downtown Chicago for the same thing.

 

[sindows]

The problem with ATI is that normal "consumers" still believe they have horrible driver support. For some reason or another, people believe that Nvidia has better drivers. IMO they're pretty equal but I do like ATI's better because they give you more options to fool around with.

 

[Rangoric]

I'd like to point out that people who play WoW alot and other games not so much if at all have no need for XFire.

WoW has both larger servers and easier communication on those servers, so the need to communicate with friends is taken care of. If anything they will have Teamspeak or equiv up and running for voice chat.

 

[Rangoric]

Originally posted by: gramboh
I don't know much about WoW, isn't it ~5k players per server and about 80 servers? So that would be 400,000 people, more than Steam yes, but still in the same ballpark. CS/HL1 games are the most popular FPS by far.

WoW is 5-6k (last I heard, although for some reason I remember reading that it can get to 10k if the players spread out well enough, which never happens) per server and there are more then 90 prolly more then 100 servers just in the United States and Oceana alone. This does not include Europe and Asia. (Edit: Europe has 84 English Servers, 64 German and 27 French Servers)

5-6 million subscribers is a total of currently active accounts, as in being paid for.

 

[gramboh]

What do you guys think about the two companies marketing? I think Nvidia does a better job and that is why they are doing better.

I'm too lazy now (should be working hehe) but I wonder if the financial statements of the two companies have segmented reporting by video card vs. integrated chipset, then you could tell who sold more.

 

[ArchAngel777]

*YAWN*... This sandbox isn't big enough for us anymore.

 

[Topweasel]

The problem with ATI is that normal "consumers" still believe they have horrible driver support. For some reason or another, people believe that Nvidia has better drivers. IMO they're pretty equal but I do like ATI's better because they give you more options to fool around with.

I disagree from the 9000 series success most people have been using ATI video cards during their current Heyday of driver devlopment, the only driver issue I think that really exists then and now is the AIW drivers. Although the availibility of Beta drivers from Nvidia probably is considered more favorable. The CCC debate probably isn't having a big affect because most people will end up getting the standard driver instead and bypassing the need for .NET and therefore also skipping over the supposed Bloat. If you are right then its ATIs fualt for not doing enough to change peoples minds.

I'd like to point out that people who play WoW alot and other games not so much if at all have no need for XFire.

Team speak and AIM are good enough programs for communication and successful enough to make Xfire useless if communication was the only desire. Xfires value is to to have an AIM type program that will tell you what your freinds are playing, giving you the choice of whther you want to join in or do your own thing. As the numbers show alot of people spend alot of time playing WoW while using team speak so discounting the numbers completely seems like a waste of effort.

What do you guys think about the two companies marketing? I think Nvidia does a better job and that is why they are doing better.

I think this is the important part. Nvidia is much better at marketing (not neccissarily Viral). I think some people just feel better from choosing from a known quantity when the performance is close. Not saying ATI isn't a safer pick, but just about everything in the US has been built up by brand Loyalism. Ford, GM, and DC spent the 80's and 90's resting on their lorels because of this, as long as nothing horrible happens people didn't change their purchasing habits and even then after they have developed a loyalism they are willing to get bit more often then not.

I have a friend who is a perfect example, he wants to build build a new machine soon. I suggest going with a 939 setup and he passes. Why? because he feels that Intels performance in games is close enough to AMDs and even at a higher price for similar performance its worth it to not take a chance. I tell him that intel has the Conroe coming up that looks good. His responce? Good! now he deffinitely doesn't need to purchase AMD products he says, he goes on about AMD saying that he likes AMD and what they are doing, by keeping themselves cometitive it has lowered the costs of Intel CPUs then says but besides that why would I purchase anything but Intel, they have always been good to me. This is after three years of talking about the growing holes between bang for the buck and performance.

For alot of people (myself slightly included) were burned with earlier ATI purchases and all of the none Riva Nvidia products were considered top notch. One blown gen from Nvidia isn't going to erase 6 years of success. Where as ATIs Foundation started severly cracked (Crummy drivers tied in with less then stellar rage cards) and I really think it will take huge screwups by Nvidia to semi perminently lose alot of their high and mid range customers. ATI can only chip away with better performance (neligible on any gen) and Visual Quality (up for debate but ATI usually leads) Features (up till recently ATI has always been a gen behind SM3 DX8 and S3TC comes to mind) and timely releases (good as Nvidia till the 7800GTX vs. X1800 both have stepped up greatly (with 7800GTX 512 being step back for Nvidia). The chipping was going nice (see X*** series) but the X1800 delay may have cost them big time.

 

[v8envy]

Originally posted by: sindows
The problem with ATI is that normal "consumers" still believe they have horrible driver support. For some reason or another, people believe that Nvidia has better drivers. IMO they're pretty equal but I do like ATI's better because they give you more options to fool around with.

ATI drivers were great during the 9800's heyday, but are slowly degenerating. Yes, Omega repackaged ATI drivers are in the same ballpark as NVidia's drivers, but the ATI distributed CCC encumbered drivers, not so much.

And if you factor Linux into the equation you would quickly determine that ATI doesn't believe drivers are anywhere near as important as NVidia.

edit: fixed typo

 

[hemmy]

Originally posted by: Topweasel

Originally posted by: rstrohkirch
Someone mentioned it above, this is a STEAM survey - not a HL2 engine survey

Go to the steam website and look at the monthly online player totals for their games and you will see just how many cs 1.6/hl1 based players there are

Wow I should have looked at that sooner. Dang 3x as many players playing CS then users playing CS:S. Wow that suprises me as well. Guess that explains alot of the 5200s, Geforce4s , and X300 users.

Why would anyone play that instead of CS:S is beyond me, even though I knew it would still be large portion of useit I didn't think it would be that much of a difference or even that CS would even be in the lead.

i have a good computer and play 1.6 over source simply because 1.6 is fun and source is not

 

[BenSkywalker]

Quite worng on the Xfire thing, I have several friends who only play RTSs ad MMOs and they run Xfire all the time. I am about the only consistent FPS player in that little group and I am not a big fan of using Xfire, to me its jus away for them to tell I am on and calling me a wuss for not playing the MMO they are playing at the time or allow them to go on and on about the size of the group they are in.

Really? Not a single MMORPG player I know uses XFire at all- what point is there? For CS it is handy if you are keeping up with your clan. Besides that, XFire is entirely optional for players of the games- having hardware is NOT optional for those on Steam. That is why it isn't comparable in terms of using it for data. If you were to polly only those Americans that drove Bugattis about the economy I'm sure you would get quite a different perspective then if you only polled homeless people. I know those aren't directly comparable- but it is akin to polling DVR users and trying to figure out general TV watching habits from them when they are in a miniscule minority and have quite different habits then the general populace.

I think this is the important part. Nvidia is much better at marketing (not neccissarily Viral).

If you don't think ATi does viral marketing then you are very naive. Intel and AMD do it too along with almost every tech company that has any interest in doing well for themselves and which there can be enthusiasts for.

 

[Matthias99]

Originally posted by: BenSkywalker

Um, HL/CS players are included in this survey. WoW has in the area of 5-6 million subscribers, and even the numbers you gave showed just HL selling, what, 10 million copies? That doesn't seem like it is 'dwarfed' to me. I don't feel like I am going to get far trying to convince you.

And how does that in any way relate to current CS players which was very explicitly what we were talking about? You explain it to me, because I'm not seeing it.

HL/CS players are included in the survey. You said:

Even when limiting it to FPSs, CS is dwarfed by the original HL.

But anyone playing the original HL or CS is included in these numbers. I don't know what you're getting at, basically.

The x700 was explicitly released as the direct competitor to the 6600GT- it was in price point and it was released after it. It is very cut and dry.

The x800/x800GTO also both had direct competitors- the 6800 and 6800GS respectively, that were very clearly defined and not a hint of the companies trying to hide that fact. In fact, they were quite clear on exactly where they were positioned.

I think you're taking the corporate positioning too literally in the face of how the cards actually performed and were priced. The X800GTO is NOT a card that competes directly with the 6800GS/6800GT in any real sense; the X800XL/XT and X850Pro were in that segment.

I'm not seeing the midrange NVIDIA dominance that was being claimed earlier.

In simple market terms, for anything close to current gen it comes down to this-

NVIDIA GeForce 6600 Series 78,439

ATI Radeon X600 11,889
ATI Radeon X700 Series 16,215

That is the straighforward both IHV endorsed midrange segment. Even if you include all of the x800 parts(we'll call it 41K) the 6600 series still comes out ahead. That is the midrange dominance people were talking about earlier. It isn't remotely close.

IMO, you can't include the 6600GT as 'midrange' and disregard the X800/X800GTO entirely; it is clearly the closest competitor to the card on the ATI side in terms of both price and performance. The X800 was between the 6600GT and 6800NU; the X700Pro was between the 6600 and 6600GT. I'll have to see if I can dig up some historical pricing information, but it is hard to find.

Yes, NVIDIA sold a lot more 6600s and 6600GTs than ATI sold X600/X700s. But in that same time frame, you could also get an X800SE/X800/X800GTO, or a RADEON 9800/9800Pro (in AGP), and people were buying them. You can't just ignore those sales if you want to look at the whole "NVIDIA versus ATI" picture.

 

[BenSkywalker]

I think you're taking the corporate positioning too literally in the face of how the cards actually performed and were priced. The X800GTO is NOT a card that competes directly with the 6800GS/6800GT in any real sense;

AT on the 6800GS-

This upper midrange, sub $250 part is NVIDIA's answer to the X800 GTO and the as of yet unavailable X1600 line up.

Link.

How is it that the GTO and GS don't compete directly against each other? They were at the same price point and I can assure you with absolute certainty they were what the people on this board were putting forward as the two most viable parts for quite some time. The two were much closer in market position then the 1900xtx and 7900GTX are.

IMO, you can't include the 6600GT as 'midrange' and disregard the X800/X800GTO entirely; it is clearly the closest competitor to the card on the ATI side in terms of both price and performance.

The 6800NU and GS were dollar for dollar matched against the x800 and GTO respectively. How is it that the 6600GT was closer?

You can't just ignore those sales if you want to look at the whole "NVIDIA versus ATI" picture.

As I pointed out- even if you include ALL of the x800 series nVidia still comes out ahead, easily, with just the 6600 family. It is by x0% instead of x00%, but nV still dominates no matter how you look at it.

 

[Matthias99]

Originally posted by: BenSkywalker

I think you're taking the corporate positioning too literally in the face of how the cards actually performed and were priced. The X800GTO is NOT a card that competes directly with the 6800GS/6800GT in any real sense;

AT on the 6800GS-

This upper midrange, sub $250 part is NVIDIA's answer to the X800 GTO and the as of yet unavailable X1600 line up.

Link.

How is it that the GTO and GS don't compete directly against each other? They were at the same price point and I can assure you with absolute certainty they were what the people on this board were putting forward as the two most viable parts for quite some time. The two were much closer in market position then the 1900xtx and 7900GTX are.

You're talking very recent history. This is part of the problem -- these cards shifted from almost high-end to what is now 'low midrange' over the last two years, and I don't think you're taking that into account very much at all.

IMO, you can't include the 6600GT as 'midrange' and disregard the X800/X800GTO entirely; it is clearly the closest competitor to the card on the ATI side in terms of both price and performance.

The 6800NU and GS were dollar for dollar matched against the x800 and GTO respectively. How is it that the 6600GT was closer?

The 6800GS launched a LOT later than these cards, and it was NOT the same price as an X800/X800GTO at its launch (or if it was, the X800/X800GTO quickly became cheaper).

You can't just ignore those sales if you want to look at the whole "NVIDIA versus ATI" picture.

As I pointed out- even if you include ALL of the x800 series nVidia still comes out ahead, easily, with just the 6600 family. It is by x0% instead of x00%, but nV still dominates no matter how you look at it.

You've lost me. I have no idea where you are getting these numbers from.

 

[BenSkywalker]

You're talking very recent history. This is part of the problem -- these cards shifted from almost high-end to what is now 'low midrange' over the last two years, and I don't think you're taking that into account very much at all.

Of course two years ago hardware that was nearly a year away from being released would be very expensive.

The 6800GS launched a LOT later than these cards, and it was NOT the same price as an X800/X800GTO at its launch (or if it was, the X800/X800GTO quickly became cheaper).

I quoted Anand's article, the 6800GS was explicitly released targetting the GTO and was released at the same price point. That is how it actually happened. I provided links- do the same.

You've lost me. I have no idea where you are getting these numbers from.

Don't trust my source? It's YOU. I'm quoting the numbers you posted.

NVIDIA GeForce 6600 Series 78,439

ATI Radeon X600 11,889
ATI Radeon X700 Series 16,215
ATI Radeon X800 Series 40,302

ATi- 68,406
nV 78,439

For a difference of 15%. Maybe my problem is I don't have an ATi approved calculator? Or are you saying I shouldn't trust your numbers?

 

[hans007]

Originally posted by: Matthias99

Originally posted by: BenSkywalker

To console gaming, yes. In comparison to the 'niche' of 'enthusiast PC gamers', IMO no.

WoW dwarfs it also- and that is never scheduled to hit consoles. Even when limiting it to FPSs, CS is dwarfed by the original HL. Online FPSs is a very small niche even when looking at PC gaming.

Um, HL/CS players are included in this survey. WoW has in the area of 5-6 million subscribers, and even the numbers you gave showed just HL selling, what, 10 million copies? That doesn't seem like it is 'dwarfed' to me. I don't feel like I am going to get far trying to convince you.

In all honesty- how do you expect anyone to make any sense out of your appraisal of the situation when you include the x700- which came out after the 6600 series and cost as much as the GT varriant? Because the GT was so far superior to what ATi was offering at that price point? Is that the logic behind it? I'm going to ignore the hands down most popular part from this survey because it doesn't agree with what I want to say. Statistical analysis can show you how to twist the numbers to say what you want them to, but even the most convuluted method of twisting wouldn't allow what you are doing.

Gee, Ben, tell me how you really feel. Could you try posting with a little more melodrama?

It's very, very hard from these numbers to break things down nicely like you want to, because there is nothing here that indicates when the cards were bought and what price was paid for them. And it seems you disagree with my interpretation. Let me try to explain:

IMO, the X600/700 basically competed (well) with the FX5500/5600/5700 and (poorly) with the GF6600, and the X800/X800GTO competed with the 6600GT/6800NU (although there was a gap in time where ATI basically had no direct competitor in that price segment). But the X800 numbers are lumped in with the X800XL/XT, and the 6800NUs are in with the 6800GT/Ultras.

For reference, here are the numbers I'm looking at:

NVIDIA GeForce4 Series 21,008
NVIDIA GeForce FX 5500 15,973
NVIDIA GeForce FX 5600 11,806
NVIDIA GeForce FX 5700 Series 17,319
NVIDIA GeForce 6200 Series 13,630
NVIDIA GeForce 6600 Series 78,439
NVIDIA GeForce 6800 Series 47,413

ATI Radeon 9500/9700 9,489
ATI Radeon 9550 13,650
ATI Radeon 9600 Series 56,484
ATI Radeon 9800 Series 43,801
ATI Radeon X600 11,889
ATI Radeon X700 Series 16,215
ATI Radeon X800 Series 40,302

That covers pretty much the entirety of the mid- to high-end of the previous hardware generation. If you total them up, you get:

NVIDIA: ~205,000
ATI: ~191,000

But what I don't know is the breakdown of the "X800" and "6800" buckets -- the 6800NU and X800[GTO] could maybe reasonably be considered "midrange" cards (we're not using the same definitions, apparently), but the X800XL/XT and 6800GT/Ultra are not. To me, it seemed like there was a definite step up in both price and performance between the 6600/X700 and the 6600GT/6800NU/X800 -- enough that they should not be considered together. Maybe that wasn't the best way to analyze it, but you don't have to rip me and accuse me of trying to twist the numbers to get the result I wanted.

It seems pretty evenly split to me no matter how you slice it; I'm not seeing the midrange NVIDIA dominance that was being claimed earlier. ATI got behind quite markedly at the higher end after the launch of the 7800GT, but the X800/X800GTO has been a pretty solid midrange alternative to the 6600GT since its launch.

i believe they bucket them like that because the driver doesnt differentiate between different x800s and 6800s.

when you install the driver it just says geforce 6800 series, and i think that is all the info that steam gets sent.

 

[Topweasel]

Originally posted by: BenSkywalker

Quite worng on the Xfire thing, I have several friends who only play RTSs ad MMOs and they run Xfire all the time. I am about the only consistent FPS player in that little group and I am not a big fan of using Xfire, to me its jus away for them to tell I am on and calling me a wuss for not playing the MMO they are playing at the time or allow them to go on and on about the size of the group they are in.

Really? Not a single MMORPG player I know uses XFire at all- what point is there? For CS it is handy if you are keeping up with your clan. Besides that, XFire is entirely optional for players of the games- having hardware is NOT optional for those on Steam. That is why it isn't comparable in terms of using it for data. If you were to polly only those Americans that drove Bugattis about the economy I'm sure you would get quite a different perspective then if you only polled homeless people. I know those aren't directly comparable- but it is akin to polling DVR users and trying to figure out general TV watching habits from them when they are in a miniscule minority and have quite different habits then the general populace.

I think this is the important part. Nvidia is much better at marketing (not neccissarily Viral).

If you don't think ATi does viral marketing then you are very naive. Intel and AMD do it too along with almost every tech company that has any interest in doing well for themselves and which there can be enthusiasts for.

First things first, I know ATI does viral marketing. No I don't have proof but it would be a fatal mistake if they didn't. Normal advertising is really good for keep your name and new releases in a persons mind but recruiting new users starts by word of mouth and other forms of viral marketing (which word of mouth is so remember whether your part of a focus group or not every good thing you say about a company is part of viral marketing). I was just noting that Nvidia has better marketing then ATI, does more work with developers (TWIMFDSWTF) not just viral marketing which I tried to cut off before someone jumped on them for it.

Also I am not saying this is the end all be all of information from Xfire, but both of us have reletively microscopic groups of friends and therefore no basis on what Actual Xfire usage is. The fact that Both CS:S and WoW dwarf any other programs on the list, even if CS:S is scewed due to more usefullness with Xfire, Does show that Valve/steam games old a large portion of use. Which this in turn shows that the survey is pooled from a large group of gamers that the stats hold value.

I think the Valve survey holds the most wieght of the two but the Xfire stats serves to validate the value of the numbers.

ATi- 68,406
nV 78,439

For a difference of 15%. Maybe my problem is I don't have an ATi approved calculator? Or are you saying I shouldn't trust your numbers?

How can you guys count the X800 numbers without counting the 6800 series. X800XL, X800XT, and X800XT PE are all high end cards that would be counted against the lowly 6600 where as many 6800s the Nu and GS come to mind that were meant to compete at the $200 or lower (eg mid-range) price segment.

 

[poohbear]

Originally posted by: Topweasel

The truths from this (and trust me a 500,000 sample group is a great sample)?

1. It looks like the 6000 series is the the current modern champion.
2. No matter what over all sales numbers that ATI and Nvidia have thrown out for all to see it is very obvious that ATI has taken a heavy hit on high sales since the introduction of the X800 and not the X1800 as most would think.
3. Middle end cards have been even worse for ATI, Nvidia out numbers the 9600 by over 20,000 and the next middle end card to do well is the X700 at 17,000 total.
4. Some would think the numbers get distorted by the ages of some of the cards, but over all with the Geforce4 MX and Geforce 4 cards still listed I don't think that holds much wieght.

To look at the numbers yourself go here.

erm, i dont think u know how surveys work. is this a random sample? it relies on ppl actually owning HL2 just to participate, which excludes probably 60% of actual users (assuming 40%, which is a very liberal estimate, of computer users actually have HL2). how often can u submit your results? are our previous results removed when u submit your new results? i.e. would i still be listed under 512mb of ram even if i upgraded to 1gb ram? HL2 was released in Nov 2004, so would these results be going back to systems ppl owned in 2004? i owned a athlon XP in 2004, now i own an opty 144, so does valve update that info or is each of my settings included as a separate stat on their survey? if not, that would explain why there are 700k samples, it might be only be 200k users but each one changed 1 item in their hardware over the past 2 years, so they count for 2/3/4 samples.