[TPU] Did NVIDIA Originally Intend to Call GTX 680 as GTX 670 Ti?

Very disappointing generation.

Really wish the 7970 was a better card so we wouldn't be stuck with Nvidia's mainstream GPU at $500.

I should have purchased the GTX 570 a year ago instead of holding out for this mess.

I realize it's just the reality of business, but this generation simply came off as a lazy effort from each side. They're both complacent and realize you have no other choice, so they are barely competing. For some reason AMD is competing with last generation cards in both price and performance, while Nvidia is just doing the bare minimum to beat AMD. Not a great path we're heading down. Enjoy the race to the bottom in performance with the same pricing.
 
Last edited:
Very disappointing generation.

Really wish the 7970 was a better card so we wouldn't be stuck with Nvidia's mainstream GPU at $500.

I should have purchased the GTX 570 a year ago instead of holding out for this mess.

I realize it's just the reality of business, but this generation simply came off as a lazy effort from each side. They're both complacent and realize you have no other choice, so they are barely competing. For some reason AMD is competing with last generation cards in both price and performance, while Nvidia is just doing the bare minimum to beat AMD. Not a great path we're heading down. Enjoy the race to the bottom in performance with the same pricing.

At least on Nvidia's side, they may have been overly cautious to not repeat the Fermi problems with moving to a new process and then just got lucky that AMD was too conservative with their clocks. If AMD had clocked the cards higher (and it isn't really clear why they didn't) Nvidia wouldn't have had the luxury of pricing the GK104 so high.
 
Only reason I bitch so much about the price is because my budget doesn't permit this type of luxury video card but I'd love to have one.

If you could afford more expensive stuff you wouldn't be as unique or recognizable on these forums!
 
Drivers will only make these better same as any other GPU but what I'm seeing here and abroad makes me come to this conclusion:

If the 680s aren't "true flagship" cards and they're already kicking as much ass as they are even here in the early goings with first run drivers...I can't wait to see what the GK110s are going to be like. That makes me think that AMD could be in serious trouble.

I'd like to believe we're coming to a place with the technology where maybe we can have flagship GPU's that don't necessarily have to be lawnmowers and small nuclear devices but I guess until we see the GK110s the jury is still technically out.
 
At the end of the day if YOU don't like the price/name/whatever of the card then don't buy it.

If I were running Nvidia I would do the same, at the end of the day, when running a publicly traded company my main obligation is to make the share holders happy, and what do share holders like? Profit.

Your personal morals do not matter to Nvidia, if you don't buy the card, someone else will.

The other issue is we don't have all the facts, just tidbits of information, and plenty of speculation. Nvidia may of very well intended the GK110 to be the flagshp, but there could be major production issues that cost $$$, or maybe producing the chips in a reasonable quantity needed to produce videocards was not cost effective, and it would of cost much more money to produce. Maybe they decided to use the GK104 as their flagship in the end after months of possible GK110 delays, so they can get something out (please shareholders), and then continue to refine the GK110. This is also speculation.
 
Just because something is initially designed to fit a certain category, doesn't mean that it doesn't end up in another. If Nvidia had a faster product they would of released it and charged $800 for it.
 
You mean like nvidia first released GTX 480 and then later GTX 580.

There is a major design difference between GK104 and GK110. The GTX 480 was literally a GTX 580 with a disabled cluster.


Just because something is initially designed to fit a certain category, doesn't mean that it doesn't end up in another. If Nvidia had a faster product they would of released it and charged $800 for it.

Why would they release another GTX 480 space heater when they have something that can compete while they finish respinning their enthusiast part for a good product instead of a rushed product.
 
There is a major design difference between GK104 and GK110. The GTX 480 was literally a GTX 580 with a disabled cluster.




Why would they release another GTX 480 space heater when they have something that can compete while they finish respinning their enthusiast part for a good product instead of a rushed product.

Because you don't spend a huge amount of money on R&D for a product then not release it because something else you own competes, look at the GTX480, they could of held onto it and respun it till they had the 580 ready to ship but you just don't do that if you have a WORKING product. Most people don't care that it sucks up 100more watts as long as it performs.
 
Because you don't spend a huge amount of money on R&D for a product then not release it because something else you own competes, look at the GTX480, they could of held onto it and respun it till they had the 580 ready to ship but you just don't do that if you have a WORKING product. Most people don't care that it sucks up 100more watts as long as it performs.

The difference being of course the fact that the GTX 480 didn't have a mid range counterpart that could compete with AMD's high end. They had to release the 480. Even if it was too early.

They don't have to release GK110 until it is ready. I don't think you understand the yield issues you run into with a die size that big.
 
The difference being of course the fact that the GTX 480 didn't have a mid range counterpart that could compete with AMD's high end. They had to release the 480. Even if it was too early.

They don't have to release GK110 until it is ready. I don't think you understand the yield issues you run into with a die size that big.

Agreed. nV bought time with the super efficient GK104. The GK110 would have been another Fermi Flamethrower in it's current state.
But I wonder if this will push back their next arch if they name it the GTX 780. A GTX 880 refresh will have to follow or it could be one and done.
It would be interesting to see if they keep it in the 600 series, and then refresh it for the 700 series.
 
Agreed. nV bought time with the super efficient GK104. The GK110 would have been another Fermi Flamethrower in it's current state.
But I wonder if this will push back their next arch if they name it the GTX 780. A GTX 880 refresh will have to follow or it could be one and done.
It would be interesting to see if they keep it in the 600 series, and then refresh it for the 700 series.

It does seem to be a bit of name gamesmanship with Nvidia trying to catch their first digit up to AMD so the uninformed won't assume the higher number is better. Microsoft did the same thing with the 360 because they didn't want to have Xbox 2 competing with Playstation 3.
 
Just because something is initially designed to fit a certain category, doesn't mean that it doesn't end up in another. If Nvidia had a faster product they would of released it and charged $800 for it.

I have to agree with that. Didn't they do that with the 8800 GTX Ultra? Even if they had the fastest chip. Sitting on tech advances is always a bad idea... information gets out...

nvidia seemed to always have scheduled a highend chip for the late summer/fall, next gen.

Would it have made everyone feel better if AMD had named their high end 6970XT & nvidia 580 TI Ultra instead of 7970 & 680, but charged the same price?

/shrug, names are just marketing...
 
I shall reiterate: They could have called it Butterfly7000superamazingrenderingmachine the second, as long as it performed the same for this price I would have bought it.

If they had called it the 670ti and been crystal clear the 680GTX would come later I still would have bought it.
 
The difference being of course the fact that the GTX 480 didn't have a mid range counterpart that could compete with AMD's high end. They had to release the 480. Even if it was too early.

They don't have to release GK110 until it is ready. I don't think you understand the yield issues you run into with a die size that big.

You are missing it, you are claiming to know that this was the mid range product, infact no one other then Nvidia knows if that is true or not. Everyone is claiming this as fact when it is not.
 
You are missing it, you are claiming to know that this was the mid range product, infact no one other then Nvidia knows if that is true or not. Everyone is claiming this as fact when it is not.

No one may know for sure, but the evidence is pretty compelling. It is pretty clear there is a bigger Kepler chip out there (it was referenced in several GTX 680 reviews), there rumored naming changes, the core naming (104 has been mid-range before), the early rumored price leaks, the bus width, the power connectors, the die size. Nvidia's comments about the 7970s performance - all those things point toward this being a mid-range part that they were able to re-brand as a high-end part.
 
You are missing it, you are claiming to know that this was the mid range product, infact no one other then Nvidia knows if that is true or not. Everyone is claiming this as fact when it is not.

Thank you Trimlock I argued this same thing in the gtx 680 leak thread. it's so nice to see the occasional person on these forums with a modicum of sanity.

The GTX 680 is a great card and I think anyone that's not a raving AMD fanboy will admit that it's the better card given it's near identical and in most cases slightly better performance and lower price. But only a raving Nvidia fanboy would call the 680 a mid range card; $500 dollar cards are not mid range unless you are Bill gates or Scrooge Mac Duck.

Does Nvidia Have another card in the works of course they do (News flash somewhere in the depths of RD they probably are working on stuff we won't see for 2-3 more years even) So does AMD but something else coming down the road in 6 months to a year does not change anything in the market today.
 
No one may know for sure, but the evidence is pretty compelling. It is pretty clear there is a bigger Kepler chip out there (it was referenced in several GTX 680 reviews), there rumored naming changes, the core naming (104 has been mid-range before), the early rumored price leaks, the bus width, the power connectors, the die size. Nvidia's comments about the 7970s performance - all those things point toward this being a mid-range part that they were able to re-brand as a high-end part.

And again, there may be a bigger core in plans or being worked on, but if its not being produced or working theres no reason to suggest this isn't the GTX680 (mainly because it is what Nvidia is calling it). Honestly thats not such a bad thing to assume this had plans to be a mid range part prior. Facts are Facts, its the best GPU Nvidia has out on the moment and they generally reserve the GTX-X80 name for that one, does it not fit?

My issue is with people saying Nvidia is holding out on a part (thats amazing people still think this) just to "one up" its competition and only because they are being competitive on a cheaper product. Thats just poor business practices.
 
This whole generation is a fucking mess, I don't care how anyone tries to spin it. We've got amd going full fucking retard with their pricing, and nvidia releasing their mid tier card as a top tier card because their real top dog is probably too fucked up to release yet. For all intents and purposes, the performance landscape looks just like it did with the last generation with nv's 570 trading blows with the 6970. The only difference is that we don't have a 580 from nv to be the big dog but we got it's price, and amd priced their shit like it was gold plated pussy. I said it before and I'll say it again, this generation is a complete clusterfuck, I'm holding onto my money until these two companies get their shit together. /rant
 
My issue is with people saying Nvidia is holding out on a part (thats amazing people still think this) just to "one up" its competition and only because they are being competitive on a cheaper product. Thats just poor business practices.

I don't think they are holding out a part, I think they had a top-end part designed that they weren't able to produce for whatever reason (yield, power use, etc) and so they re-branded the mid-range part as the high-end one while they got their act together on the big part. When they first conceived of the Kepler line-up a year ago, or whenever, I really doubt this is the chip they had planned to be the top-end part. They just got a little bit lucky that they were able to make it work that way.
 
It does seem to be a bit of name gamesmanship with Nvidia trying to catch their first digit up to AMD so the uninformed won't assume the higher number is better. Microsoft did the same thing with the 360 because they didn't want to have Xbox 2 competing with Playstation 3.

It more likely it will become the GTX 780 since it will release in or after August.
We should see a dual GK104 (GTX 690) to fill the gap until Q4.

It has to be sickening for AMD to know nV has a chip that will beat the 8900s refresh months before they hit the market.

Thank you Trimlock I argued this same thing in the gtx 680 leak thread. it's so nice to see the occasional person on these forums with a modicum of sanity.

Just because you agree with his opinion doesn't mean you two have factual info. The GK100 was rumored since last year and dumped in January due to heat and yield issues.
The GK110 was developed in it's place but it would take months for it to tape out. The GK104s on the on the other hand were also rumored last year and ready to go. Smaller chips are easier to produce.
Look at the spec of the 7870 and then look at the GTX 680 and say with a serious face the GTX 680 is the flagship card. The Kepler arch is just much more powerful than AMDs.

This isn't info that just popped up on the net two weeks ago. There's a timeline of leaks dating back to November.
 
It more likely it will become the GTX 780 since it will release in or after August.
We should see a dual GK104 (GTX 690) to fill the gap until Q4.

It has to be sickening for AMD to know nV has a chip that will beat the 8900s refresh months before they hit the market.



Just because you agree with his opinion doesn't mean you two have factual info. The GK100 was rumored since last year and dumped in January due to heat and yield issues.
The GK110 was developed in it's place but it would take months for it to tape out. The GK104s on the on the other hand were also rumored last year and ready to go. Smaller chips are easier to produce.
Look at the spec of the 7870 and then look at the GTX 680 and say with a serious face the GTX 680 is the flagship card. The Kepler arch is just much more powerful than AMDs.

This isn't info that just popped up on the net two weeks ago. There's a timeline of leaks dating back to November.

But you are the one that does not have proof all you have are "leaks and rumors" no one except Nvidia knows what their original release plan was for all you know gk 110 turned out to be a dead end because every time you turned on a PC with the card inside it caught on fire blew up and kicked a baby on it's way out.

I on the other hand do have proof I have 100's of benchmarks and reviews from reputable sites showing that the gtx 680 is in most instances the fastest single GPU in the world and that it is named like a high end product and priced like a high end product.

Saying product x is not high end because something else might come out in 6 months that's faster is ridiculous was the gtx 480 not a high end gpu because the 580 came out ?
is the 7970 not a high end gpu because the 680 exists maybe by you definition but not by and generally accepted one.

I can see a duck and say no that's not really a duck because real ducks shoot lazer beams from their eyes and grant wishes and if that's my definition of what a real duck is then I would be correct but only to myself, everyone else would think I was nuts.
 
I on the other hand do have proof I have 100's of benchmarks and reviews from reputable sites showing that the gtx 680 is in most instances the fastest single GPU in the world and that it is named like a high end product and priced like a high end product.

Saying product x is not high end because something else might come out in 6 months that's faster is ridiculous was the gtx 480 not a high end gpu because the 580 came out ?
is the 7970 not a high end gpu because the 680 exists maybe by you definition but not by and generally accepted one.

No one is questioning that the GTX 680 (GK104) is Nvidia's high-end product now - it obviously is. The discussion was over whether it was intended to be the high-end part all along, or if it was put in the high-end position for some other reason. I think the evidence points pretty clearly to it being put into the high-end role, and not originally designed for it.
 
No one is questioning that the GTX 680 (GK104) is Nvidia's high-end product now - it obviously is. The discussion was over whether it was intended to be the high-end part all along, or if it was put in the high-end position for some other reason. I think the evidence points pretty clearly to it being put into the high-end role, and not originally designed for it.

It could be, but again saying clearly is a bit misleading since we do not have any facts just assumptions.

I'll say it again I have no problem with people wanting to state Nvidia had plans prior for it to be midrange but then stating "Nvidia is holding out", "the only reason they haven't released the true GTX 680 is lack of competition."

I'm sorry but it seems most people want this wet dream that there is something better out there and there simply isn't. Just because they had plans originally to release a higher end chip that doesn't mean anything as they have not. This reminds me of the 3dfx specter or whateveritsnameis. The one thing that was supposed to demolish nvidia but never got released or somesuch.
 
But you are the one that does not have proof all you have are "leaks and rumors" no one except Nvidia knows what their original release plan was for all you know gk 110 turned out to be a dead end because every time you turned on a PC with the card inside it caught on fire blew up and kicked a baby on it's way out.

I on the other hand do have proof I have 100's of benchmarks and reviews from reputable sites showing that the gtx 680 is in most instances the fastest single GPU in the world and that it is named like a high end product and priced like a high end product.

Saying product x is not high end because something else might come out in 6 months that's faster is ridiculous was the gtx 480 not a high end gpu because the 580 came out ?
is the 7970 not a high end gpu because the 680 exists maybe by you definition but not by and generally accepted one.

I can see a duck and say no that's not really a duck because real ducks shoot lazer beams from their eyes and grant wishes and if that's my definition of what a real duck is then I would be correct but only to myself, everyone else would think I was nuts.


Pretty much, its a conspiracy theory to state that another product exists and is being manufactured ... just waiting ... to pounce the competition!
 
the only thing that confuses me about the GK104 is the 128 texture mapping units. If this was designed for the mid range wouldn't it have closer to 64 or 80 similar to AMD's midrange cards?

If this was intended to be the midrange they couldn't have been clueless about the performance this would bring to the midrange.

Yes I realize TMUs are one small component of a system but the GK104 has 128 and Tahiti XT has 128 seems odd.
 
Too many people hung up on the bit size of the memory bus, this is their "hammer" to their logic. What they don't understand is the amount of BW it still offers is very damn high.

TMU's are most likely required because of TXAA, they could of cut down on quite a bit of AA performance and saved even more power (even die size) if they cut down the TMU's but it seems that Nvidia is banking a lot on TXAA.
 
But you are the one that does not have proof all you have are "leaks and rumors" no one except Nvidia knows what their original release plan was for all you know gk 110 turned out to be a dead end because every time you turned on a PC with the card inside it caught on fire blew up and kicked a baby on it's way out.

I on the other hand do have proof I have 100's of benchmarks and reviews from reputable sites showing that the gtx 680 is in most instances the fastest single GPU in the world and that it is named like a high end product and priced like a high end product.

Saying product x is not high end because something else might come out in 6 months that's faster is ridiculous was the gtx 480 not a high end gpu because the 580 came out ?
is the 7970 not a high end gpu because the 680 exists maybe by you definition but not by and generally accepted one.

I can see a duck and say no that's not really a duck because real ducks shoot lazer beams from their eyes and grant wishes and if that's my definition of what a real duck is then I would be correct but only to myself, everyone else would think I was nuts.

Did this topic make you upset?

I own a GTX 680, so isn't not like I'm speaking negatively about nV and the GK104.
Just bringing up a good question if we got the intended part, and a reputable site with ties in the industry thought it was worth making an article about.

They'll name it the GTX 780, and that will validate your argument....well sort of.
Look at the [H] review, they mentioned the GK110 at the end in their article. Also Kyle many times spoke of a part not being ready until the Summer.
I guess the GK104 was a conspiracy in January/February too. Rumors are eventually verified or discredited and we haven't reached either conclusion on the GK110.
 
Doesn't matter what they have in the future, if it ain't here it doesn't exist in consumer world. So get off the fanboy horse and call it what it is. It is a 500 dollar card and it is enthusiast card.

I have nvidia in my system, I have had ati in my system, and I will go back and forth in the future. So stop bitching about a card that is no where to be found and judge by what is out in the market.

so what if it was suppose to be called gtx 670 ti, it is called gtx 680 and it is priced like a gtx 680. It is fast as hell and I am sure AMD is getting their shit ready for the gtx 780 if nvidia calls it that.

It has happened before and it will happen again.
 
Doesn't matter what they have in the future, if it ain't here it doesn't exist in consumer world. So get off the fanboy horse and call it what it is. It is a 500 dollar card and it is enthusiast card.

I have nvidia in my system, I have had ati in my system, and I will go back and forth in the future. So stop bitching about a card that is no where to be found and judge by what is out in the market.

so what if it was suppose to be called gtx 670 ti, it is called gtx 680 and it is priced like a gtx 680. It is fast as hell and I am sure AMD is getting their shit ready for the gtx 780 if nvidia calls it that.

It has happened before and it will happen again.

u mad?
 
And here comes the GTX 685 :p http://www.expreview.com/18901.html

2012032602353927.jpg


Edit: I'm not trying to start anything, I just haven't heard about this before...
 
Yeah, my bad Forceman. I actually just found that same chart somewhere from 4 days ago. That being said, I'm going to back out of this conversation, and go back to waiting for 680s to be in stock again so I can buy one lol
 
Last edited:
104 looks better with the lower power and heat.

what chip will be used for the gtx 660?

my gtx 460 is still easily good enough for my needs.
 
Back
Top