1060 vs 1070 FPS/$

[agentagent47]

I want to know the FPS/$ of each cards compared to the other. 

[manikyath]

fps/$ is kind of a dumb metric, because intel integrated graphics are the winner.

[Ichi]

1 minute ago, manikyath said:

fps/$ is kind of a dumb metric, because intel integrated graphics are the winner.

That's why he's asking about specific card instead of the general "what is the best".

No harm done in answering that.

[Prindagelf]

1 minute ago, manikyath said:

fps/$ is kind of a dumb metric, because intel integrated graphics are the winner.

i do not even want to know what rock you have been living under..

 

but as for this question, it will greatly depend on the brand and ultimate clocking along with what you can find as far as actual pricing goes.

[manikyath]

Just now, Ichi said:

That's why he's asking about specific card instead of the general "what is the best".

No harm done in answering that.

my point was it makes no sense.

 

would you get a card half the speed of the best one you can afford because the fps per dollar is slightly higher?

[manikyath]

1 minute ago, Prindagelf said:

i do not even want to know what rock you have been living under..

ellaborate?

[MaxBunny]

12 minutes ago, agentagent47 said:

I want to know the FPS/$ of each cards compared to the other. 

You should look at youtube benchmarks for all of the games you want to see the FPS of with both to decide that. But I would expect the 1060 to have a better value per frame. The 70 and 80 are premium cards with bad value ranges in general.

[MageTank]

17 minutes ago, manikyath said:

fps/$ is kind of a dumb metric, because intel integrated graphics are the winner.

Price to performance is the only metric that should matter to most consumers. As for the prospect of "Intel wins", that is simply not true. Too many variables at play. Run a 4k game on max settings, and the Intel iGPU will show you a stop-motion picture film. Run a game on the absolute minimum settings, and a high end dGPU will have hundreds to thousands of FPS. 

 

You knew very well what his question was. Do not treat it as if its a dumb question. If you cannot answer it, simply move on.

[manikyath]

Just now, MageTank said:

Price to performance is the only metric that should matter to most consumers. As for the prospect of "Intel wins", that is simply not true. Too many variables at play. Run a 4k game on max settings, and the Intel iGPU will show you a stop-motion picture film. Run a game on the absolute minimum settings, and a high end dGPU will have hundreds to thousands of FPS. 

 

You knew very well what his question was. Do not treat it as if its a dumb question. If you cannot answer it, simply move on.

the thing about intel graphics, is they're there the moment you have a modern cpu

(sorry amd FX guys, you know i'm right.)

 

and people dont care about fps per dollar directly, they care for value for their money. but if your monitor is 1440p 144Hz, and you already have a GTX780, the "best fps per dollar" card is actually not very good value, because it's not a very big improvement, and it cannot push the monitor to its fullest just like the old solution could.

 

alike, if OP had a more mid range cpu (lets say an i5 4460) with no near future plans to upgrade, and generally playing more cpu bound games (lets say cities:skylines) 

 

the question OP should ask is what is the better card for him to get, based on his needs, his budget, and the system it is gonna drop into.

 

fps per dollar is "just one metric" that does not cover the entire story, which if detached from other metrics (like the actual produced fps) produces a false truth.

[MageTank]

54 minutes ago, manikyath said:

the thing about intel graphics, is they're there the moment you have a modern cpu

(sorry amd FX guys, you know i'm right.)

 

and people dont care about fps per dollar directly, they care for value for their money. but if your monitor is 1440p 144Hz, and you already have a GTX780, the "best fps per dollar" card is actually not very good value, because it's not a very big improvement, and it cannot push the monitor to its fullest just like the old solution could.

 

alike, if OP had a more mid range cpu (lets say an i5 4460) with no near future plans to upgrade, and generally playing more cpu bound games (lets say cities:skylines) 

 

the question OP should ask is what is the better card for him to get, based on his needs, his budget, and the system it is gonna drop into.

 

fps per dollar is "just one metric" that does not cover the entire story, which if detached from other metrics (like the actual produced fps) produces a false truth.

Then you agree that there are variables at play that change the outcome. However, nothing changes when you compare how strong a piece of hardware is, vs what it costs compared to another similarly priced piece of hardware. In this case, the GTX 1060 vs GTX 1070. If a card costs 30% more, but only performs 10% better, it has worse performance per dollar, regardless of your monitor. However, you are right in the sense that someone could be looking at specific GPU's, and have a CPU bottleneck. That is very true. 

 

The fact remains that his question could be answered without all of these semantics added to the mix. You can answer his question AND educate him on potential bottlenecks without calling the metric dumb, and then providing false information saying the Intel iGPU's are "the winner". The Iris Pro SKU's alone already go against that theory, as those specific SKU's cost upwards of $100-$150 more than their normal Intel HD counterparts. They've yet to make one that competes with the 750 Ti, an $80-$120 GPU. 

 

Price:Performance depends on the time of day (as prices of hardware changes) and is independent of other pieces of hardware. When comparing two pieces of hardware specifically, bottlenecks are assumed to be non-existent. 

[BrightCandle]

You can guarantee that FPS/$ will always be higher with the lower card than the higher one. There are quite a few reasons for that but the primary one is to do with the silicon process and bigger die sizes have lower yields and as such cost more. You also tend to have higher end VRAM and a wider bus on the bigger card to attempt to feed it sufficiently. Power circuitry goes along with bigger dies as well as they consume more power and its just "wasted" money in that its necessary to make the card run but doesn't contribute to the performance of the card directly.

 

So the lower end cards, whether harvested (1070) or smaller die (1060) are going to be better FPS/$ than the cards above them, and that will be true for the 1050 over the 1060 as well.

 

The only case this isn't true is with the really low end GPUs where 3D performance isn't the goal. The cost of those isn't driven by the die and its support but the cost of an actual card and so their performance/$ is worse than the card directly above it for 3D gaming, but they are mainly for business machines for extra monitor plugs.

[agentagent47]

ok. everyone seems to be triggered and didn't answer my question  

[joeydutchman]

Search on Google next time. You'll have your answer faster