GTX1070TI spotted

Opps yeah you are correct on the dates.

No 1080ti sales will not be hurt anyways. 1080ti is a 700 card, 1% of the performance bracket in sales, whom ever bought those cards would have bought them already. It will raise opportunity cost, but the is minor, inventory numbers won't get hurt, if they are managing their manufacturing volume vs sales volume, last I checked nV's inventory figures have been fairly stable ;)

Steam survey disagrees with your assessment of 1080 Ti sales curve.

http://store.steampowered.com/hwsurvey/videocard/

Sales of 1080 Ti are steady over time, as compared to the massively trailed off sales of the 1080 (which only recently got a bump in sales due to the 1060 and 1070 being marked up massively due to mining)

Steam Survey DX12 GPUs:

NVIDIA GeForce GTX 1080 Ti
- April
- May
0.42% June
0.53% July
0.66% August
+0.13%

NVIDIA GeForce GTX 1080
1.65% April
1.64% May
1.70% June
1.74% July
1.98% August
+0.24%

As compared to the 1080, which was always priced at a premium before the 1080 Ti was released (and still had somewhat of a premium afterwards before the mining spiked prices of 1060 and 1070), the 1080 Ti >>is<< the value choice for GP102, which means that it's sales are guaranteed for the life of the product, as there are always people upgrading to the highest performance value product over the course of the life of such a product (Until you reach too close to the expected launch of the next generation of course).
 
Steam survey disagrees with your assessment of 1080 Ti sales curve.

http://store.steampowered.com/hwsurvey/videocard/

Sales of 1080 Ti are steady over time, as compared to the massively trailed off sales of the 1080 (which only recently got a bump in sales due to the 1060 and 1070 being marked up massively due to mining)

Steam Survey DX12 GPUs:

NVIDIA GeForce GTX 1080 Ti
- April
- May
0.42% June
0.53% July
0.66% August
+0.13%

NVIDIA GeForce GTX 1080
1.65% April
1.64% May
1.70% June
1.74% July
1.98% August
+0.24%

As compared to the 1080, which was always priced at a premium before the 1080 Ti was released (and still had somewhat of a premium afterwards before the mining spiked prices of 1060 and 1070), the 1080 Ti >>is<< the value choice for GP102, which means that it's sales are guaranteed for the life of the product, as there are always people upgrading to the highest performance value product over the course of the life of such a product (Until you reach too close to the expected launch of the next generation of course).


And how does that sustained sales, slow down with Pascal's release? Volume sales go up and and level off again as before then :

Opportunity cost only happens when you lose the opportunity to make certain amount of money, releasing a new product to replace the other product at the same price points or lower and maintaining margins, Yes maintaining margins because 12nm is a cheaper process than 16nm. how do you explain opportunity cost when nV will have increased margins?

BTW this is the 1070 sales figures

3.36%
3.30%
3.55%
3.41%
3.85%

Sure mining is stopping for selling 1070's, I think nV starting making more 1070's after the mining craze started

They haven't leveled off, still going up. lets take this a step further all Pascal nV cards went up last month why?
Vega launch was disappointing.

1060 went up 1.26% that is higher than the 1070.

Again stream is great to look at trends, but don't use it to show people that its a definitive answer.
 
Last edited:
Not sure how people are disagreeing with my captain obvious statements.....

Of course people will continue purchasing 1080 Ti (the value GP102) until it starts getting close to time for Volta launch.

The longer that Value GP102 is on the market the more sales you are going to get, this is simply obvious.

Top value chip for each silicon, 1070 for GP104, 1080 Ti for GP102, will have steady sales effectively forever as long as a successor/replacement hasn't comes along yet, thats just how this kind of market works.

People don't buy 50-90% of the total volume of the life of the product of a value product immediately.
 
Not sure how people are disagreeing with my captain obvious statements.....

Of course people will continue purchasing 1080 Ti (the value GP102) until it starts getting close to time for Volta launch.

The longer that Value GP102 is on the market the more sales you are going to get, this is simply obvious.

Top value chip for each silicon, 1070 for GP104, 1080 Ti for GP102, will have steady sales effectively forever as long as a successor/replacement hasn't comes along yet, thats just how this kind of market works.

People don't buy 50-90% of the total volume of the life of the product of a value product immediately.


Doesn't matter, when they can get a cheaper product with the same performance or higher with similar or higher margins. Your opportunity cost is no longer exists. That is captain obvious :).

Also high end graphics cards, need to be refreshed at a certain time, they just don't keep selling. These aren't CPU's where nV or AMD can rely on businesses to take care of the revolving door. If they don't innovate in a timely manner they will lose potential sales. You need to look at the money makers here, that 4% 1070 and 3% 1080 is like 50 times more volume than that 1080ti sales. (JPR figures). So where is the opportunity cost going to be the lose of performance sales or the potential lose of the 1080ti sales?

When you upgrade your cards tell me, is it because games come out and you want to play them, with higher fidelity? When do business upgrade their CPU's? Most business have a 4-5 year turn around on systems, its in their IT manuals or rule book, what ever they want to call it, they must do it. GPU's don't have that type of sustained sales.
 
Spoon feed failure.

I give up, you guys can keep living in your personal bubbles.

Don't mind me, I've only predicted every single launch since GTX 480 correctly.

Clearly I know nothing......
 
Spoon feed failure.

I give up, you guys can keep living in your personal bubbles.

Don't mind me, I've only predicted every single launch since GTX 480 correctly.

Clearly I know nothing......


Well I've been doing it since geforce fx and only got one thing wrong that was the 48xx series, performance, everything else hit it perfectly.

Pretty much what you are saying it goes against Nvidia's 1.5 years per gen launch schedule. Which they haven't wavered from since Fermi. They talked about it after the g80 they will go to that, and they did which since then its been pretty much rock solid.
 
Well I've been doing it since geforce fx and only got one thing wrong that was the 48xx series, performance, everything else hit it perfectly.

Pretty much what you are saying it goes against Nvidia's 1.5 years per gen launch schedule. Which they haven't wavered from since Fermi. They talked about it after the g80 they will go to that, and they did which since then its been pretty much rock solid.

Tell me what you believe I believe, you don't seem to be making sense in any of your arguments since none of them contradict mine, and then you claim that they make mine false.
 
Tell me what you believe I believe, you don't seem to be making sense in any of your arguments since none of them contradict mine, and then you claim that they make mine false.


Well opportunity cost is not there, as I explained its money they could get but don't get it, if that product is replaced and they get the same amount of money from the new product, then there is no opportunity cost.

I'll give you an example, if I was to buy a 580 mining rig right now for eth mining vs a 1070 mining rig, the opportunity cost of the the 580 rig at this point is there, cause I'm not going to make my money back. for the 1070 I can switch over to other alt coins and still get that money back at the same rate (actually higher) than the 580's. Hence why I sold most of my 580 rigs already.
 
hUNAo.jpg
 
Well opportunity cost is not there, as I explained its money they could get but don't get it, if that product is replaced and they get the same amount of money from the new product, then there is no opportunity cost.

I'll give you an example, if I was to buy a 580 mining rig right now for eth mining vs a 1070 mining rig, the opportunity cost of the the 580 rig at this point is there, cause I'm not going to make my money back. for the 1070 I can switch over to other alt coins and still get that money back at the same rate (actually higher) than the 580's. Hence why I sold most of my 580 rigs already.

I'm not going to explain basic concepts like opportunity cost here, my spoonfeed level is already at the maximum I can tolerate.

You believe what you want.
 
I'm not going to explain basic concepts like opportunity cost here, my spoonfeed level is already at the maximum I can tolerate.

You believe what you want.

http://examples.yourdictionary.com/opportunity-cost-examples.html

I'm not sure what you are talking about, but that's not opportunity cost. Opportunity cost is the potential loss of something. If there is no loss there is no opportunity cost. Replacing an existing product, there is never any opportunity cost. What you are talking about is left over inventory (write off) or R&D not being fully recovered or profits realized. R&D is a fixed cost when talking about tech and must reoccur, so it doesn't have an opportunity cost, since it must still go on. Pretty much it comes down to is non realized profits is what you are getting at. Unrealized profits are never looked at unless a company is going to a different direction than their previous product.

So tell me how do you have opportunity cost, when the product is replaced, and you still make the same money? There is none right?

Now lets say this. Individual margins are less on the performance cards vs the 1080ti. But the volume they sell at is much higher.

Now if we start looking at price, demand, supply curves over the different market areas, such as OEM's vs Consumer market, and factor in inventory left over. Opportunity cost will be there some where right? Cause the new product will cause a negative demand on the older product, but you need to factor in for how long do they have inventory for. This is why I mentioned, nV doesn't seem to have any issues with inventory management right now. So can't even look at that.

Lets talk about tech, CUDA, what was nV's opportunity cost with CUDA? Well what would it have been if AI was never there?

It would have been all the money they put into it to create an SDK and dev relations right?

Lets look at opportunity cost of HBM?

The cost of creating the product, cost of creating the channels, because for AMD its not their product so for them its a risk. For Hynix it was a risk as well. Its new to the market and must be adopted, that is its opportunity cost, adoption rate.

The actually product, Fiji graphics card, there is no opportunity cost, since it came to market and replaced well superseded the previous high end products. The rest of the development and manufacturing is pretty much the same as any other graphics card, it must be done, fixed costs for development and manufacturing (both of these behave as variable costs, but in actuality are fixed costs, because there is no way around spending that money).

I'll give you an example of when nV miscalculated inventory levels. Pascal's release, and the 970. It caused much more left over 970's (they didn't expect this because of the increased price of the 1070's, the 970's were in the lower level of the performance cards, the 1070's were price in the mid level of performance segment), but they didn't need to write it off, they just sold them to OEM's at full price.

Price break down of the segments

mainstream 0-150
mid range 150-300
performance 300 to 500
enthusiast 500+

Same thing AMD did with Fiji sell to OEM's to get rid of as much stock as possible at full price just several months ago.. So no opportunity costs.
 
Last edited:
There is news about prices of graphics card rising even further, minimum another 10% due to memory shortages and memory price increasing. So good luck finding 1070ti for 400. They are all going up in price it seems.
 
There is news about prices of graphics card rising even further, minimum another 10% due to memory shortages and memory price increasing. So good luck finding 1070ti for 400. They are all going up in price it seems.


Problem is all graphics cards prices will be higher than MSRP till mid next year because of the ram price hike.
 
There is news about prices of graphics card rising even further, minimum another 10% due to memory shortages and memory price increasing. So good luck finding 1070ti for 400. They are all going up in price it seems.

The option to not buy is always available as a choice ...
 
Last edited:
nah they won't announce gaming cards in a tech conference, its probably just more talk about Volta in HPC and DL space.
 
Drive down 1070 closer to MSRP. In turn, drive down 1060 closer to MSRP.


... is it too much to dream?
 
Likely 479 ish.
How so, when GTX 1070 currently sell new for $469+ ??? and 1080 is $530+? I'm not seeing where the 1070 Ti price fits in to the mix. There's a price hole between $600 (highest end 1080) and $730 (lowest end 1080 Ti). No way most folks will pay $600+ for a 1070 Ti, would just buy a 1080. Or am I seeing the pricing scheme all wrong?
 
How so, when GTX 1070 currently sell new for $469+ ??? and 1080 is $530+? I'm not seeing where the 1070 Ti price fits in to the mix. There's a price hole between $600 (highest end 1080) and $730 (lowest end 1080 Ti). No way most folks will pay $600+ for a 1070 Ti, would just buy a 1080. Or am I seeing the pricing scheme all wrong?

They are saying it retails for 429.99. I was just adding minimum 50 to it. yea I don't know and it boggles my mind why nvidia is giving importance to vega 54. Its just a dumb move to be honest. it's like giving attention to your competition when there is no competition. May be this is another special card for miners
 
release date was posted in this thread as Oct 26 so Afterburner might be a good confirmation that release will be this month ?
 
release date was posted in this thread as Oct 26 so Afterburner might be a good confirmation that release will be this month ?

Supposedly that's the "paper launch" date, with in-store availability in the first week of November.
 
im currently thinking Ill just rush out and order one as soon as possible. Figure the longer I wait the more expensive everything will get right now. If I can get one for $420 Ill take it.
 
im currently thinking Ill just rush out and order one as soon as possible. Figure the longer I wait the more expensive everything will get right now. If I can get one for $420 Ill take it.

Considering you can't get a regular 1070 for that price, no idea how the Ti's are going to be available there. I mean, huzzah if they are, because that means that 1070's should be back in the mid-high 300's.
 
Hmmm... here's two theories:
  • Leftover 1080 GP104's that didn't quite make the cut + more base RAM than 8GB
  • Higher clocked 1070 GP104 + keep the RAM at 8GB but use GDDR5X
Either way, this is an easy way to make a new product out of existing parts, take on Vega, conquer the "mid-range," etc., etc., etc.
 
Hmmm... here's two theories:
  • Leftover 1080 GP104's that didn't quite make the cut + more base RAM than 8GB
  • Higher clocked 1070 GP104 + keep the RAM at 8GB but use GDDR5X
Either way, this is an easy way to make a new product out of existing parts, take on Vega, conquer the "mid-range," etc., etc., etc.
Why on earth would they add more vram? There is nothing logical about that at all as the next step up on the 256 bit bus would be 16 GB. It is just basic common sense that it will be 8 GB of vram just like the 1070 and 1080.
 
Man I hope they have Blower Style Coolers available. Those are usually only on reference cards and supposedly Nvidia isnt making a reference design for this.
 
Back
Top