NVIDIA GeForce GTX 275 review [Guru3D.com]

[GabrielSans]

Originally Posted by Duke Nil

Yeah lol, it really is kind of absurd that people are complaining about nvidia offering more and better products... The fact that they release a new card doesn't mean that your card is any worse and it doesn't cost you anything. People are talking as if nvidia told them to their face that they have to upgrade if they want to continue playing video games, that they will be fined and insulted if they don't stay on the absolute bleeding edge of the market.

Wrong! People here are not complaining about variety of AMD ATI videocards but Nvidia ones. And they are not complaining about the variety itself but the way it was created: Nvidia has been in a "rebranding/enabling-disabling shader processors" spree too long!

They went wrong from the very start, IMHO, when they released GTX 280 and GTX 260 on a big 65 nm die. If you look at those two cards, the only "big" difference you find is the amount of shader processors enabled: 240 and 192 respectively. In short, it is the "same" core with some tweaking.

On the other hand, the release of HD Radeon 4870 and HD Radeon 4850 was much more consistent in my opinion. First of all, the dice is 55 nm. Both cores are identical but with different clock speeds. Anyway, the big difference is that 4870 uses GDDR 5 while 4850 uses GDDR 3. My point is that these different memories justify AMD ATI is naming those two videocards in a different way. The prices are also justified because the difference in performance is very noticeable.

Except shrinking the die down to 55 nm (which they should've done with the GTX 280 and 260 from the very word "Go"), I just can't see much work on the Nvidia's part. For example, when they released the absurd GTX 260 SP216, apart from shrinking the die (as I told, something they should've done from the start), they just enabled more shader processors. The chip and memory (even their clock speeds) are exactly the same as always!

Not to mention that the 448 bits bus generating things like 896 MB and 1792 MB of memory is really weird to say the very least. All this just adds more confusion to the mostly already confused buyers.

Well, let's sum it up: I am not a Nvidia basher (in fact, I bought/used their videocards since 2000 through 2008). I'd like this company to come back to a wise course of action in releasing their videocards. What they did with the 200 series is literally too much "of just tweaking a little some things". And unfortunately, if they will continue this absurd course of action throughout 2009 we'll see more ridiculous hybrids and rebranding without any serious improvement. Their attitude of running after any AMD ATI videcard being released is simply childish in my view.

All in all, I'd like to see Nvidia returning to their good old days, period.

[Terrorizor]

Originally Posted by GabrielSans

All in all, I'd like to see Nvidia returning to their good old days, period.

100% agree . geforce ti series comes to mind.

[snip3r_3]

I still find it weird why people think AMD's products are so well named, when...

9800GTX+/GTX250 to GTX260 performance difference is the same as a 4850 to a 4870 ...
Just like 4870 -> 4890 = GTX260 -> GTX275
Similar?


Regarding the just more shaders, and shrink to 55nm. Doing a die shrink is not as easy as you think it is. Its not just shrink, it requires plenty of considerations and is complicated, as well as money-consuming. If it was that easy, why don't think you they released a better version from the start?

I really think that people should be more objective in viewing both companies. What AMD/Nvidia are doing is exactly the same. Well slightly different, but overall, the same thing.

GDDR3/GDDR5 difference would equate to Nvidia's SP differences. Clock speed differences exist on both sides. It doesn't matter which of these are changed, but the end result is... There are levels in the product generation, each one performing better than the next. What one manufacture does, for example change clock speed, may work better on one GPU architecture than the other. There really no reason for saying that one company's ways for differentiating their products is bad, as both ways work and in the end, thats all that counts.

To everyone stating that the 275 is just slightly faster than the 260 C216, it is (as it is marketed as a GTX275, not greater than 280/285, thus is in between the two cards)! However, then is the 4890 such a giant leap over the 4870? If you actually look between the reviews here at guru3d, then you will notice that... in fact, the jump from a C216 to a 275 is actually greater (as by frame rates) compared to the difference between a 4870 to a 4890 (non-oc). Therefore, didn't both companies do the same exact thing? Both cards thus are just a tiny bump in existing product family, nothing extraordinary as they are intended to be refreshes.

Geforce Ti, whatever other series that came before, all used the same techniques for differentiating product classes. Clock speed differences, pipeline differences, and memory configuration differences all have existed to separate chips to high/medium/low end classifications.

448bit -> 896mb ram on the 260/275 VS the 512bit -> 1024mb ram on the 280/285...
=
GDDR3 256bit -> 512mb on the 4850 vs the GDDR5 256bit 512/1024mb on the 4870/4890...

Why is that confusing when it is exactly the same (both step up as the higher the class)? Nvidia uses more bus width and RAM, and AMD uses better memory technology and more ram as well... Both achieve the same goal, just in a slightly different way.

While I agree with you (and everyone) that the 8/9 series naming was a disaster (and that the C216 for the GTX260 was confusing, I believe it should have been called 265), the rest makes sense. If we look only at the current generation of cards, then the lineups from both companies make perfect sense. In fact, each product step is about the same.

GTS240 = 4830
GTS250 = 4850
GTX260 = 4870
GTX275 = 4890

Looks clean and simple doesn't it? While there are multiple ram configurations for various cards here and there, overall, everything makes sense. I'll leave out the 280/285/X2s/295 as those don't really have a direct counterpart.
280<285<4850X2<4870X2<295 with consideration that CF/SLI works perfectly in the sense of scaling performance.

I just find that it is weird why everyone is so angry at nvidia's products, while both product releases accomplished the same goals. Both offer great performance at a cheap price. The sole difference is that since AMD did not previously have a higher single-gpu card, the 4890 is the new flagship card. On Nvidia's side, since the 4890 was a step inbetween the 260 and 285, then it is logical to fill one in there as a new price segment/performance level has been established. If the 4890 was faster than the 285, then you would be sure to see a GTX290 or 289 or whatever. This round of releases was done well, as cheap cards with great performance is always welcome. Just that I believe people need to judge products using a neutral point of view.

Anyways, both cards are great, the 275/4890 is really bringing the price down on high-end performance. Just that I find no need to keep stating that Nvidia is doing the wrong thing this and that while if we take a closer look, is pretty much exactly the same thing as what AMD is doing.

----
short bullet point list to skip my confusing post

-> FPS change from GTX260 -> 275 = 4870 -> 4890
-> GDDR3/5 change + clk speed change is = SP change + memory config change + clock speed change no specific way is better than the other, both companies have achieved making different product classes
-> GTS240-GTS250-GTX260-275 = 4830-4850-4870-4890 naming/performance/price (depends on where you live)
-> 8/9 series naming sucks as well as the GTX260 c216, but that has nothing to do with the 275, whose naming is perfectly fine (260<275<285)

-> my Vista 64bit installation died mysteriously, well not really, but I got some weird message saying Recycle Bin was corrupt and then it did a check disk that replaced security things, and now its dead -_- (off-topic, but I'm angry , which makes me want to type)
-> my spring break started =] yay

[Rochin]

Originally Posted by snip3r_3

I still find it weird why people think AMD's products are so well named, when...

9800GTX+/GTX250 to GTX260 performance difference is the same as a 4850 to a 4870 ...
Just like 4870 -> 4890 = GTX260 -> GTX275
Similar?


Regarding the just more shaders, and shrink to 55nm. Doing a die shrink is not as easy as you think it is. Its not just shrink, it requires plenty of considerations and is complicated, as well as money-consuming. If it was that easy, why don't think you they released a better version from the start?

I really think that people should be more objective in viewing both companies. What AMD/Nvidia are doing is exactly the same. Well slightly different, but overall, the same thing.

GDDR3/GDDR5 difference would equate to Nvidia's SP differences. Clock speed differences exist on both sides. It doesn't matter which of these are changed, but the end result is... There are levels in the product generation, each one performing better than the next. What one manufacture does, for example change clock speed, may work better on one GPU architecture than the other. There really no reason for saying that one company's ways for differentiating their products is bad, as both ways work and in the end, thats all that counts.

To everyone stating that the 275 is just slightly faster than the 260 C216, it is (as it is marketed as a GTX275, not greater than 280/285, thus is in between the two cards)! However, then is the 4890 such a giant leap over the 4870? If you actually look between the reviews here at guru3d, then you will notice that... in fact, the jump from a C216 to a 275 is actually greater (as by frame rates) compared to the difference between a 4870 to a 4890 (non-oc). Therefore, didn't both companies do the same exact thing? Both cards thus are just a tiny bump in existing product family, nothing extraordinary as they are intended to be refreshes.

Geforce Ti, whatever other series that came before, all used the same techniques for differentiating product classes. Clock speed differences, pipeline differences, and memory configuration differences all have existed to separate chips to high/medium/low end classifications.

448bit -> 896mb ram on the 260/275 VS the 512bit -> 1024mb ram on the 280/285...
=
GDDR3 256bit -> 512mb on the 4850 vs the GDDR5 256bit 512/1024mb on the 4870/4890...

Why is that confusing when it is exactly the same (both step up as the higher the class)? Nvidia uses more bus width and RAM, and AMD uses better memory technology and more ram as well... Both achieve the same goal, just in a slightly different way.

While I agree with you (and everyone) that the 8/9 series naming was a disaster (and that the C216 for the GTX260 was confusing, I believe it should have been called 265), the rest makes sense. If we look only at the current generation of cards, then the lineups from both companies make perfect sense. In fact, each product step is about the same.

GTS240 = 4830
GTS250 = 4850
GTX260 = 4870
GTX275 = 4890

Looks clean and simple doesn't it? While there are multiple ram configurations for various cards here and there, overall, everything makes sense. I'll leave out the 280/285/X2s/295 as those don't really have a direct counterpart.
280<285<4850X2<4870X2<295 with consideration that CF/SLI works perfectly in the sense of scaling performance.

I just find that it is weird why everyone is so angry at nvidia's products, while both product releases accomplished the same goals. Both offer great performance at a cheap price. The sole difference is that since AMD did not previously have a higher single-gpu card, the 4890 is the new flagship card. On Nvidia's side, since the 4890 was a step inbetween the 260 and 285, then it is logical to fill one in there as a new price segment/performance level has been established. If the 4890 was faster than the 285, then you would be sure to see a GTX290 or 289 or whatever. This round of releases was done well, as cheap cards with great performance is always welcome. Just that I believe people need to judge products using a neutral point of view.

Anyways, both cards are great, the 275/4890 is really bringing the price down on high-end performance. Just that I find no need to keep stating that Nvidia is doing the wrong thing this and that while if we take a closer look, is pretty much exactly the same thing as what AMD is doing.

----
short bullet point list to skip my confusing post

-> FPS change from GTX260 -> 275 = 4870 -> 4890
-> GDDR3/5 change + clk speed change is = SP change + memory config change + clock speed change no specific way is better than the other, both companies have achieved making different product classes
-> GTS240-GTS250-GTX260-275 = 4830-4850-4870-4890 naming/performance/price (depends on where you live)
-> 8/9 series naming sucks as well as the GTX260 c216, but that has nothing to do with the 275, whose naming is perfectly fine (260<275<285)

-> my Vista 64bit installation died mysteriously, well not really, but I got some weird message saying Recycle Bin was corrupt and then it did a check disk that replaced security things, and now its dead -_- (off-topic, but I'm angry , which makes me want to type)
-> my spring break started =] yay

Very well spoken, and I agree 100%

[Denial]

The issue for me is upgrading from lower cards. If I have a 8800GT I'm going to expect pretty large upgrade going two series up to a GTX240, yet in fact it's the same card.

As far as putting cards in-between I don't think it's an issue at all. As long as they are named based on performance in a clear way.

A GTX265, 216 shouldn't have been called that. It should have been like a GTX270. The 55nm, if slightly faster, should be a GTX273.

[LedHed]

It's very simple why both the 4890 and GTX 275 were released, to compete at a price point where no cards are currently priced ($250). The GTX 260 is running well under $200 and the GTX 285 continues to hover around $320, creating a large gap between ~$170-$320. It's only natural for ATI and NVIDIA to pursue this market and I can't understand for the life of me why some Gurus are becoming hostile about the cards..

$250 used to be the hottest pricing point for budget minded enthusiast and NVIDIA and ATI want to open this segment up again.

Does no one remember the 8800GT/GTS 512 launch? The 8800GT launched at $250 and it competed against the 8800 GTX which was running for $450 at the time of 8800GT launch. This is almost identical to that scenario that is happening now with the GTX 275 vs 285.

I can't understand why people are complaining about a card that is much cheaper than the GTX 285 (by $100) and almost performs identical (less than 8% difference on avg).

Take a look at the performance numbers:

GTX 260: 216SPs 72 TUs 28 ROPs 448bit BUS
GTX 275: 240SPs 80 TUs 28 ROPs 448bit BUS
GTX 285: 240SPs 80 TUs 32 ROPs 512bit BUS (same as GTX 280)

It is pretty obvious why NVIDIA chose the performance numbers they did for the 275, they increased stream processors and texture address units because they are simpler to increase than the ROPs/BUS. This is the same reason ATI refuses to increase their BUS and instead use GDDR5 to make up the difference in memory bandwidth, because an increased BUS is much more expensive in overall board design than increasing VRAM speed (or TUs and SPs).

I for one think the GTX 275 is perfectly priced and named, it's the only card that NVIDIA has which falls between $200 and $300 and the logical progression of the 200 Series naming scheme leads us to GTX 275. I can even show you posts before the 275 was even announced where I was calling the 295 a hybrid "275 SLI setup". The 270 shows us it has higher performance numbers (SPs and TUs) than the 260 and the 5 tells us that it is 55nm.

I mean look at these numbers between the 285 and the 275 with 185 drivers:

http://www.bit-tech.net/hardware/gra...orce-gtx-275/1

Bit-Tech obviously feels the 285 is no longer economically viable due to the small performance difference but very large price increase (some 285's are still running for $400 before MIR (eVGA FTW))

NVIDIA has to bring down the price of the 285 or it will not be competitive with the new lineup.


Overall I think ATI and NVIDIA have both done a fine job with this refresh, giving 10-20% performance increase over the cards they replaced with the same price point is what all of us want to see. The only way people could be upset is that their expensive cards are now being toppled by mid-range cards.

[snip3r_3]

Originally Posted by Denial

The issue for me is upgrading from lower cards. If I have a 8800GT I'm going to expect pretty large upgrade going two series up to a GTX240, yet in fact it's the same card.

As far as putting cards in-between I don't think it's an issue at all. As long as they are named based on performance in a clear way.

A GTX265, 216 shouldn't have been called that. It should have been like a GTX270. The 55nm, if slightly faster, should be a GTX273.

True, but a 8800GT is a high-end card, technically, since its x800. Uhh, like though there is the GTS GTX Ultra, it still belongs to the enthusiast segment of the 8 Series.
The GTS240/250 are the low end part of the GTX/S family, so naturally, the performance will be around last generation's high end cards (which is spot on XD)

If you gave the average person, a 9800GTX+ and a GTS250 (1gb), then he would be happy to state that the GTS250 is a great new low-end card since it is cheap, offers the same performance as previous generation's high-end card, and consumes less power and has more RAM. 512mb version of the GTS250, would have the same performance, but it costs less than the 1gb version, so it is reasonable.

This trend exists in every generation jump, such as a FX5900 -> 6600, 6800->7600, each new mid/low end card offered the performance of the previous generation's high-end cards.

If Nvidia were to develop a new card to fit below the GTX260, the performance would still be equal to the 9800GTX+ which would be superfluous as there was a product that does just that already. Though it is a bit of a lazy thing to do instead of developing a true mid-range/low-end product, the decision to directly rename/slightly redesign the GTS250 means they can lower the price further as compared to a new chip based off the GT200.

Now G92 renamed 8 series to 9 series really is retarded, I can not think of any logical explanation other than it was a market maneuver.

edit: just saw LedHed's post, which was great.
Graphs show what I previously stated, 260->275 has greater performance change than 4870->4890
I also noticed how close the 275 and 285 were... o_o extremely close...

[alanm]

Debates over 2 closely competing cards (ie, 4890 vs GTX 275) can be misleading. Anyone can cherry pick games that favor one card over the other to give a convincing impression of one cards superiority over the other. I would never take a single review of one card as definitive (even if they were accurate reviews from reputable sources), I would have to see at least 3 or 4 reviews under varying conditions with different games to get a better idea. Plus one new cards performance today may be different (better) after weeks and months of new driver releases, and heres where Nvidia come out on top imo: driver optimizations.

p.s. I'm not critiquing this review in any way, only the debates between 2 cards (as we've seen above). Guru3d reviews are my first preference in getting an informative take on any graphics card, thats for sure

[Aura89]

Originally Posted by GabrielSans

They went wrong from the very start, IMHO, when they released GTX 280 and GTX 260 on a big 65 nm die. If you look at those two cards, the only "big" difference you find is the amount of shader processors enabled: 240 and 192 respectively. In short, it is the "same" core with some tweaking.

Um...i'd say that's a good bit of difference between the cards, i mean cmon, look at the 7800 GTX vs the 7800 GT, the main differences between the two were core speeds and memory speeds and slight difference between pipelines etc, same goes to say with the 6800 Ultra and 6800 GT, since the 8800 GTX came out, there has been even bigger differences between the cards then i have seen for quite some time of the two "big" ones typically released...

I mean what else do you want? a completely different card? a card that has so much less performance then a GTX 280 yet priced the same as a GTX 260 was when it first came out? absolutely no choice other then the GTX 280? how about a GTX 280 and a GTX 260 and the GTX 260 have a core speed of 300, memory 450, etc.? I have absolutely no idea what problem you have with what currently nvidia is doing between the cards, i can completely understand the renaming, that's just getting annoying, however, what you are complaining about (where i quoted) makes no sense....

[alanm]

Also I recall the 6600gt wiped the floor with FX5900, despite being a new mid-range card vs a last gen top card. They were not close as implied by someone earlier.

[nam]

Why no 280 comparison?

[The Postman]

I would take the HD4890 any day, its overclocking potential is amazing and it is its selling point also.

That card at 1ghz leaves behind the GTX285

[GabrielSans]

Originally Posted by snip3r_3

I still find it weird why people think AMD's products are so well named, when...

9800GTX+/GTX250 to GTX260 performance difference is the same as a 4850 to a 4870 ...
Just like 4870 -> 4890 = GTX260 -> GTX275
Similar?


Regarding the just more shaders, and shrink to 55nm. Doing a die shrink is not as easy as you think it is. Its not just shrink, it requires plenty of considerations and is complicated, as well as money-consuming. If it was that easy, why don't think you they released a better version from the start?

I really think that people should be more objective in viewing both companies. What AMD/Nvidia are doing is exactly the same. Well slightly different, but overall, the same thing.

GDDR3/GDDR5 difference would equate to Nvidia's SP differences. Clock speed differences exist on both sides. It doesn't matter which of these are changed, but the end result is... There are levels in the product generation, each one performing better than the next. What one manufacture does, for example change clock speed, may work better on one GPU architecture than the other. There really no reason for saying that one company's ways for differentiating their products is bad, as both ways work and in the end, thats all that counts.

To everyone stating that the 275 is just slightly faster than the 260 C216, it is (as it is marketed as a GTX275, not greater than 280/285, thus is in between the two cards)! However, then is the 4890 such a giant leap over the 4870? If you actually look between the reviews here at guru3d, then you will notice that... in fact, the jump from a C216 to a 275 is actually greater (as by frame rates) compared to the difference between a 4870 to a 4890 (non-oc). Therefore, didn't both companies do the same exact thing? Both cards thus are just a tiny bump in existing product family, nothing extraordinary as they are intended to be refreshes.

Geforce Ti, whatever other series that came before, all used the same techniques for differentiating product classes. Clock speed differences, pipeline differences, and memory configuration differences all have existed to separate chips to high/medium/low end classifications.

448bit -> 896mb ram on the 260/275 VS the 512bit -> 1024mb ram on the 280/285...
=
GDDR3 256bit -> 512mb on the 4850 vs the GDDR5 256bit 512/1024mb on the 4870/4890...

Why is that confusing when it is exactly the same (both step up as the higher the class)? Nvidia uses more bus width and RAM, and AMD uses better memory technology and more ram as well... Both achieve the same goal, just in a slightly different way.

While I agree with you (and everyone) that the 8/9 series naming was a disaster (and that the C216 for the GTX260 was confusing, I believe it should have been called 265), the rest makes sense. If we look only at the current generation of cards, then the lineups from both companies make perfect sense. In fact, each product step is about the same.

GTS240 = 4830
GTS250 = 4850
GTX260 = 4870
GTX275 = 4890

Looks clean and simple doesn't it? While there are multiple ram configurations for various cards here and there, overall, everything makes sense. I'll leave out the 280/285/X2s/295 as those don't really have a direct counterpart.
280<285<4850X2<4870X2<295 with consideration that CF/SLI works perfectly in the sense of scaling performance.

I just find that it is weird why everyone is so angry at nvidia's products, while both product releases accomplished the same goals. Both offer great performance at a cheap price. The sole difference is that since AMD did not previously have a higher single-gpu card, the 4890 is the new flagship card. On Nvidia's side, since the 4890 was a step inbetween the 260 and 285, then it is logical to fill one in there as a new price segment/performance level has been established. If the 4890 was faster than the 285, then you would be sure to see a GTX290 or 289 or whatever. This round of releases was done well, as cheap cards with great performance is always welcome. Just that I believe people need to judge products using a neutral point of view.

Anyways, both cards are great, the 275/4890 is really bringing the price down on high-end performance. Just that I find no need to keep stating that Nvidia is doing the wrong thing this and that while if we take a closer look, is pretty much exactly the same thing as what AMD is doing.

----
short bullet point list to skip my confusing post

-> FPS change from GTX260 -> 275 = 4870 -> 4890
-> GDDR3/5 change + clk speed change is = SP change + memory config change + clock speed change no specific way is better than the other, both companies have achieved making different product classes
-> GTS240-GTS250-GTX260-275 = 4830-4850-4870-4890 naming/performance/price (depends on where you live)
-> 8/9 series naming sucks as well as the GTX260 c216, but that has nothing to do with the 275, whose naming is perfectly fine (260<275<285)

-> my Vista 64bit installation died mysteriously, well not really, but I got some weird message saying Recycle Bin was corrupt and then it did a check disk that replaced security things, and now its dead -_- (off-topic, but I'm angry , which makes me want to type)
-> my spring break started =] yay

I never said that shrinking to 55 nm was easy. I just said that they should've done that in the first place with GTX 260 and GTX 280. It was a failure from the start. They should have controlled their obssesion, hunger for money or whatever, and released the right die size such as AMD ATI did. On top of that, when they released GTX 260 and 280, they charged people a lot of money. From the current prices that are almost half of the original ones, it is evident that they were cheating people and making big profits in the process. This is not good, of course.

Nvidia is not doing the same as AMD, sorry. Nvidia has also cheated people with rebranding, for example. The same chip G92 has gone through three generations by now. This is a marketing cheat, no doubt about it.

Your list:

GTS240 = 4830
GTS250 = 4850
GTX260 = 4870
GTX275 = 4890

doesn't look clean and simple on the Nvidia's side really if you look at it closely. For example, those GTS and GTX prefixes are a mystery to most people, let alone they make the name much longer. 240, 250, 260 and 275 would be something simple, in my opinion.

While you find no need to keep stating that Nvidia is doing the wrong thing, some other people do (e.g. myself). This is the way things work. Your problem is that you behold everything from a Nvidia guru's viewpoint (i.e. someone that understands fully all about Nvidia products and specs), but most people are not one. Most people will find much more consistent the AMD ATI's nomenclature and this is a fact you can verify by just asking people if you wish. The inclusion of prefixes such as GT, GTS and GTX are quite a mystery to Joe User, and this is no surprise. I'd like to know how many people know exactly the difference between them.

When you say:

"-> FPS change from GTX260 -> 275 = 4870 -> 4890
-> GDDR3/5 change + clk speed change is = SP change + memory config change + clock speed change no specific way is better than the other, both companies have achieved making different product classes
-> GTS240-GTS250-GTX260-275 = 4830-4850-4870-4890 naming/performance/price (depends on where you live)"

if you read attentively your words, you'll see that you have forgot to include GTX 280. Apparently, GTX 280 is better than GTX 275 by following the nomenclature, isn't it? Anyway, I am not sure from the specs. If I, a hardcore gamer who have read so many Hilbert's articles for so long, have a real problem to say which one is the best, imagine Joe User going to buy one Nvidia card. The core of this mess is Nvidia releasing GTX 260 and 280 in a wrong way, i.e. with a 65 nm die. They were in a hurry and made a mistake. From that mistake, plus the addition of mysterious prefixes, plus rebranding and crazy SP216, they land on this paraphernalia of confusing names.

From your explanation, it all looks like Nvidia is not doing anything wrong, but this is so because you understand perfectly what's going on. Most people don't, regretfully. The AMD ATI nomenclature is much more solid and understandable, and till now with their HD 4000 series they have not cheated people with rebranding. They did it fine from the beginning and that's why I will continue to buy Radeons for now.

I repeat: I am not a Nvidia basher, but just someone pointing out obvious flaws. I cannot buy products from Nvidia anymore, while this company continues to behave in a not serious manner. It is not serious what they are doing with their 200 series, beyond the performance of the cards. Before buying hardware, apart from considering bang for bucks, I always analyze how serious a company is. Nvidia is not doing well by my seriousness standards for the time being. When they recover good judgment, possibly I will buy their products again, as I have done for so many years before.

Despite my opinions, I can understand your point, friend.

Off topic: Yes, I have Vista 64 too, and there are some weird things occurring sometimes. No registry cleaning fixes them, and the drives are OK. Well, it is part of the magic of Vista 64 . Still alive, fortunately. Anyway, I need Vista 64 because of my quantity of RAM. I keep a XP installation always, just in case.

Please, don't respond to this long post in a long way or we'll end up with quite a Bible!

[nam]

Originally Posted by The Postman

I would take the HD4890 any day, its overclocking potential is amazing and it is its selling point also.

That card at 1ghz leaves behind the GTX285

Not wearing that hat you have on profile pic.. take it off lol ...

[snip3r_3]

Originally Posted by GabrielSans

Please, don't respond to this long post in a long way or we'll end up with quite a Bible!

Well, if I'm correct, then the GTX260/280 have been discontinued, replaced by the C216 and GTX285 after their introduction.

note : Um, for me referencing to the GTX260, I pretty much mean the C216 unless I state its the old one.

Releasing it directly to 55nm would have probably caused it to be delayed. Furthermore, the power characteristics of the GTX260/280 were still acceptable, being similar to a 8800Ultra while offering much more performance. The GTX260 also had similar power characteristics to the 4870. Since the 260/280 were released before the HD4 series, it of course commanded a price premium, since it they were the fastest cards. Though the price they were released at were nothing different than Ultra/GTX cards in the past, high-end + new release = pricey.

About the prefixes, I don't think its really that misleading, as most people don't care about it, whats more significant to them would be the 3 numbers behind it. GTS/GTX -> HD is pretty much the same thing to me. If both companies would call their products in a similar manner, it would be boring (like GFX275 GFX4890). I found both companies doing the general consumer a bit better by starting to drop suffixes, the XT/PRO/GT/LE/GTS/etc really was a bigger mess for the consumers. The numbering and naming used today is quite better than it was years ago. AMD certainly has a slightly better naming structure, but I find Nvidia's still pretty clear, as numbers rise up nicely, indicating performance. Its quite like the car industry as well, where prefixes and suffixes are all over the place, every manufacturer has its own system. Furthermore, like my friend indicated when I built his computer, price sometimes informs the consumer more than the fancy numbering and words written on packaging. The pricing as well as the numbering pretty much means fast->faster->fastest to anyone. Honestly, I think that no matter what kind of numbering/naming system exists, the average consumer won't understand what they mean without research or someone to inform them.
Thus I find AMD's naming structure to be only slightly better, if at all any different from Nvidia's.

The GTX275 -> GTX280 is very narrow in terms of performance and numbering, hence, people probably would believe they performance in a similar manner (which they do). Nvidia could have made the naming worse by calling it a GTX270, though the performance difference between the GTX260 -> "GTX270" is clearly greater than the gap between the "270" and the 280.

I think that the naming is quite decent this time along, as I have mentioned, the G92 fiasco/GTX260 C216 upgrade naming should not influence the view point on this product, as the GTX275 is indeed not a rename product. (though I think the 9800GTX+ to the GTS250 is not that bad as some say it is)

-> Yes the 6600 was faster than any FX in games where the FX sucked like crap (dx9, which the FX was flawed in), but otherwise, they performed similarly. If we look at the 7600 and the 6800s, then they perform quite similarly as well. The 8600 and 7800/7900s performed similarly as well (after later driver updates), with the 8600GTS outperforming the 7900GT/GS in newer games. Since the 8/9 series were a mess-up naming wise, the next step up would be the GTS250 being similar to the 9800GTX+ (considering the new 1gb board).

Anyways, I see your point. Though I still think that both companies should have their own systems. After all, they need some sort of difference to separate their products right? :p

OT: Need to find my X64 disk to start windows repair -_- And whats worse, ants have invaded my oven (i was shocked when I tried making my breakfast)!! Thus microwave/stove cooked food for me today ...

[Terrorizor]

Originally Posted by snip3r_3

Geforce Ti, whatever other series that came before, all used the same techniques for differentiating product classes. Clock speed differences, pipeline differences, and memory configuration differences all have existed to separate chips to high/medium/low end classifications.

the point is 3-5fps difference between cards for hundreds more is retarded and the renaming is to mask that exact reason and its obvious.geforce 3 and 4 ti were far more spaced out between classes,and it was clear where they stood. so you use to be able to get the performance you paid for. not "similar " performance among them all.

the 260 is too close to the 275 which is too close to the 280 which is too close to the 285 and the previous generation is used as the mid and entry level cards so they can keep re-using cheap g92 chips with the same specs but a different name.

not to mention, how many versions of the gtx260 is there? 3 at least not including oc versions. because in that sense there was no name change at all for different cards with different shader stream amounts and manufacturing processess ,while the 285 got a new name for the same reason the gtx260 DIDNT get a name change for 55nm and the 2 different versions with different specs before it.

[alanm]

The no.1 objective of GPU manufacturers is to make money. That is NOT an unreasonable objective given that these are publicly traded companies who must answer to shareholders. I think its a bit silly for people here to second guess these companies on their product lineups, manufacturing processes, etc and offer their takes on how they should sell, name, manufacture or position these products in the market. If these companies do not maximize their profits, the shareholders will demand to replace their managements.

This is also a VERY fast moving market. Theres almost no room to breathe. I laughed at the idea mentioned above that Nvidia shouldve waited for the 55nm process before releasing the gtx260 or 280. And watch ATI in the meantime clean up that segment of the market with their 4870? Puhleez! These companies are not out to look for your best interests in the products they sell, they do what they do to keep their jobs! If I was an Nvidia shareholder I would be pissed if some ethical twit on the management board said "hey, lets wait for the 55nm process before we release this card, its in the interests of our customers, they deserve better". Or if someone said "hey, lets not name this card a 200 series number, its really a G92", it may confuse the customer into buying it". Lol!

[GabrielSans]

Originally Posted by alanm

The no.1 objective of GPU manufacturers is to make money. That is NOT an unreasonable objective given that these are publicly traded companies who must answer to shareholders. I think its a bit silly for people here to second guess these companies on their product lineups, manufacturing processes, etc and offer their takes on how they should sell, name, manufacture or position these products in the market. If these companies do not maximize their profits, the shareholders will demand to replace their managements.

This is also a VERY fast moving market. Theres almost no room to breathe. I laughed at the idea mentioned above that Nvidia shouldve waited for the 55nm process before releasing the gtx260 or 280. And watch ATI in the meantime clean up that segment of the market with their 4870? Puhleez! These companies are not out to look for your best interests in the products they sell, they do what they do to keep their jobs! If I was an Nvidia shareholder I would be pissed if some ethical twit on the management board said "hey, lets wait for the 55nm process before we release this card, its in the interests of our customers, they deserve better". Or if someone said "hey, lets not name this card a 200 series number, its really a G92", it may confuse the customer into buying it". Lol!

You have proved my point, friend! Read your statements.

So, lack of ethics regarding buyers is "permitted" in favor of getting more money and keeping the shareholders happy. Anyway, company and shareholders (and yourself too) forget one thing: the buyers are the ones supporting a company in the end.

What you have affirmed is exactly the opposite to the current marketing policies, which are centered in "client satisfaction". So, Nvidia and their shareholders (and yourself too, according to your opinions) are going in the wrong way from the viewpoint of the dominating marketing policy nowadays.

And this was my point finally: Nvidia is not behaving seriously in regard to the buyers, and that's why I won't buy their products till they recover the course "client satisfaction over quick profits/shareholders happy in a non-ethical way" again. Period.

[GabrielSans]

Originally Posted by snip3r_3

Well, if I'm correct, then the GTX260/280 have been discontinued, replaced by the C216 and GTX285 after their introduction.

note : Um, for me referencing to the GTX260, I pretty much mean the C216 unless I state its the old one.

Releasing it directly to 55nm would have probably caused it to be delayed. Furthermore, the power characteristics of the GTX260/280 were still acceptable, being similar to a 8800Ultra while offering much more performance. The GTX260 also had similar power characteristics to the 4870. Since the 260/280 were released before the HD4 series, it of course commanded a price premium, since it they were the fastest cards. Though the price they were released at were nothing different than Ultra/GTX cards in the past, high-end + new release = pricey.

About the prefixes, I don't think its really that misleading, as most people don't care about it, whats more significant to them would be the 3 numbers behind it. GTS/GTX -> HD is pretty much the same thing to me. If both companies would call their products in a similar manner, it would be boring (like GFX275 GFX4890). I found both companies doing the general consumer a bit better by starting to drop suffixes, the XT/PRO/GT/LE/GTS/etc really was a bigger mess for the consumers. The numbering and naming used today is quite better than it was years ago. AMD certainly has a slightly better naming structure, but I find Nvidia's still pretty clear, as numbers rise up nicely, indicating performance. Its quite like the car industry as well, where prefixes and suffixes are all over the place, every manufacturer has its own system. Furthermore, like my friend indicated when I built his computer, price sometimes informs the consumer more than the fancy numbering and words written on packaging. The pricing as well as the numbering pretty much means fast->faster->fastest to anyone. Honestly, I think that no matter what kind of numbering/naming system exists, the average consumer won't understand what they mean without research or someone to inform them.
Thus I find AMD's naming structure to be only slightly better, if at all any different from Nvidia's.

The GTX275 -> GTX280 is very narrow in terms of performance and numbering, hence, people probably would believe they performance in a similar manner (which they do). Nvidia could have made the naming worse by calling it a GTX270, though the performance difference between the GTX260 -> "GTX270" is clearly greater than the gap between the "270" and the 280.

I think that the naming is quite decent this time along, as I have mentioned, the G92 fiasco/GTX260 C216 upgrade naming should not influence the view point on this product, as the GTX275 is indeed not a rename product. (though I think the 9800GTX+ to the GTS250 is not that bad as some say it is)

-> Yes the 6600 was faster than any FX in games where the FX sucked like crap (dx9, which the FX was flawed in), but otherwise, they performed similarly. If we look at the 7600 and the 6800s, then they perform quite similarly as well. The 8600 and 7800/7900s performed similarly as well (after later driver updates), with the 8600GTS outperforming the 7900GT/GS in newer games. Since the 8/9 series were a mess-up naming wise, the next step up would be the GTS250 being similar to the 9800GTX+ (considering the new 1gb board).

Anyways, I see your point. Though I still think that both companies should have their own systems. After all, they need some sort of difference to separate their products right? :p

OT: Need to find my X64 disk to start windows repair -_- And whats worse, ants have invaded my oven (i was shocked when I tried making my breakfast)!! Thus microwave/stove cooked food for me today ...

Well, so Nvidia is doing nothing wrong from reading your post. OK, I and a lot of other people are completely wrong then. It was just an optical illusion. Nvidia must continue to tread this path then, as they are completely right. Now, I am relieved!

[TheHunter]

this gpu looks very tempting for its price, especially cause of those 240sp , damn why did my old gpu die so soon couldn't it wait for 1 more month

[MAD-OGRE]

I dont get why ppl are upset every time a new card comes out, its getting old, if you get so emo about it, then maybe you should pick a new hobby.
I do understand that the names changing causes some people to make poor choices, but if your going to spend more then $10.00 on some thing and not do some reserch thats your own fault.
Come on its 2009, if your buying a GPU 99% chance you have the internet and if you to dumb to do a search no one to blame but your self.

If your unhappy you spent $XXXX on a 280 or 285 to have this card come out and give you about the same preformace for a cheaper price, well take the about of days you have had your card, devide it by the price difrence and ask your self, was it worth having that card vs the card you had before for that price perday.
GTX285 $329.99
GTX275 $249.99
30days used = $80.00 difrence
so for $2.67 per day you got the power of a 285 over what you had.

ATI and Nvidia both do this, its all about money to them, and if you think it should be about any thing other then that, man what rock are you living under? I might want to join you in the land of no cares.

[alanm]

Originally Posted by GabrielSans

You have proved my point, friend! Read your statements.

So, lack of ethics regarding buyers is "permitted" in favor of getting more money and keeping the shareholders happy. Anyway, company and shareholders (and yourself too) forget one thing: the buyers are the ones supporting a company in the end.

Heh, you have this a bit twisted. I'm not really the one saying its a lack of ethics, you were the one implying it in your earlier posts (ie, Nvidias "hunger for money" etc) and which I built an argument using a hypthothetical 'false-ethics' approach which may conflict with shareholders interests. Not waiting for the 55nm and selling the 65nm gtx 260 was not in the least bit unethical, in fact if Nvidia had waited for the 55nm, it may have been unethical to their shareholders who are obliged maximum value for their stock. It could even be unethical to the customers if Nvidia rushed out the 55nm pre-maturely and found it had issues OR to those customers who had been waiting for so long to upgrade and wouldnt have minded the 65nm product. You appear to have pre-judged this case by claiming Nvidia did wrong in not waiting for the 55nm without ascertaining the facts or circumstances surrounding it.

With regard to the naming of the GTX 250, that was perfectly legit with no unethicality about it whatsoever, even if it was based on the G92 architecture. Its a mid-range product with a new name and where the model number properly suggests a relatively lower performance than the other G200 cards. Plus it was offered to the public at an amazingly low price. Kudos to them for that. This sort of thing happens often with numerous products in many areas and fields.

Originally Posted by GabrielSans

What you have affirmed is exactly the opposite to the current marketing policies, which are centered in "client satisfaction". So, Nvidia and their shareholders (and yourself too, according to your opinions) are going in the wrong way from the viewpoint of the dominating marketing policy nowadays.

"Client satisfaction" as determined by who? By you? As with all products you will hear pros and cons in many aspects and areas of any given product. The products performance and support service should be the main area of "client satisfaction", NOT on Nvidias manufacturing and marketing policies as you seem to absurdly think. Thats an area that Nvidia must deal with according to its own interests first and foremost and which naturally covers shareholders interests. Of course client satisfaction is essential to that, but you have to know where that applies for it to make any sense.

Originally Posted by GabrielSans

And this was my point finally: Nvidia is not behaving seriously in regard to the buyers, and that's why I won't buy their products till they recover the course "client satisfaction over quick profits/shareholders happy in a non-ethical way" again. Period.

Ha ha... you sound lost with your misplaced principles. Good luck on your future non-Nvidia purchases.

[GabrielSans]

Originally Posted by MAD-OGRE

...

ATI and Nvidia both do this, its all about money to them, and if you think it should be about any thing other then that, man what rock are you living under? I might want to join you in the land of no cares.

I cannot keep debating in a healthy way when someone attacks me personally. The core of this debate was Nvidia and so on, not myself. Thanks for sharing your opinions anyway. Good luck.

[GabrielSans]

Originally Posted by alanm

...

With regard to the naming of the GTX 250, that was perfectly legit with no unethicality about it whatsoever, even if it was based on the G92 architecture. Its a mid-range product with a new name and where the model number properly suggests a relatively lower performance than the other G200 cards. Plus it was offered to the public at an amazingly low price. Kudos to them for that. This sort of thing happens often with numerous products in many areas and fields.

If all this is legit, I hope they won't be rebranding the GeForce 2 GTS as a brand-new GTX 230 or something.

"Client satisfaction" as determined by who? By you? As with all products you will hear pros and cons in many aspects and areas of any given product. The products performance and support service should be the main area of "client satisfaction", NOT on Nvidias manufacturing and marketing policies as you seem to absurdly think. Thats an area that Nvidia must deal with according to its own interests first and foremost and which naturally covers shareholders interests. Of course client satisfaction is essential to that, but you have to know where that applies for it to make any sense.

I own a nForce 680i SLI that was very expensive in late 2007. I bought other two GeForce's at that time, and in previous years too. So, I think I am a customer. I am fully entitled to opine about Nvidia and their policy then.

Ha ha... you sound lost with your misplaced principles. Good luck on your future non-Nvidia purchases.

I don't like when a debate comes to a personal level. Nobody can arrive at healthy conclusions when the thing is too personal. Besides, when the things become too personal, offenses are the next step. I am not interested in that, friend.

If you are going to like everything Nvidia does, well, I won't follow your way. I liked Nvidia before and dislike it now... and maybe I will like it in the future again. I am not bound to like Nvidia all the time, and I have the right to declare that, I suppose, without offending anyone personally. If someone takes offense at my opinions about Nvidia, maybe he is one of their shareholders or simply a fan boy. No use to debate with those shareholders, fan boys, etc., as you surely know.

Good luck

[applejack]

why do ati fans keep complaining about nvidia instead of praising their own choice ?
the video cards debates should remain in the technical realm, where it actually matters to the individuals.
I never cared for all this drama and simply choose the better product for my money in terms of performances / features / software support
and other technical stuff.

all these whining and personal opinions in regards to corporate policy & monkey business behind the scenes is so irrelevant.
I would never screw my gaming experience because of some stupid principle. it is such a lame excuse for buying an inferior product
/ supporting the competitor.