Will my board support a GTX 970/80 card?

[EarthDog]

Better than a blatently and patently false statement that it's faster.

 

[P4-630]

Even though the RX480 is a 14nm GPU, it's still more power hungry than a GTX1060 6GB

 

[Ungari]

The Power Consumption, the Power Consumption! Caw Caw Rawwwwk!

 

[ASOT]

I've been a amd gpu fan since many years,had many amd gpu's models but..
when i move to 660ti<760<970 i was impress by pwm efficiency,how fast and good are,no artefacts and more damm good features physx,gsync,etc

Get nVIDIA whatever it is 970<980<Ti or Pascal 1060 6gb or better

 

[dalekdukesboy]

Yes, now there are great deals for new 980Ti's as low as $400, and the performance is close to 1080.

These are both true, actually understated I got my 980 ti couple months back for 300 bucks.

 

[EarthDog]

I've seen low 300s plenty of time...

And the 980ti is a bit slower than the 1070 gents.

 

[dalekdukesboy]

Mine was 300 period so low 300's kinda overstated in that case Also I researched before buying the ti leaning toward 1070 but was $400.

 

[P4-630]

980Ti's as low as $400, and the performance is close to 1080.

Your "close"s seem to be off a lot lately....

 

[Papahyooie]

From the research I did before getting one, the 980ti seems to be ~70% the performance of a 1080. So if you can find one for ~300 USD, it's still a heck of a deal, especially for 1080p/1440p.

EDIT: Especially if you overclock, or even get one of the factory overclocked versions. The 980ti FTW was for sale for 300 bucks for the holidays on EVGA's b-stock.

 

[dalekdukesboy]

From the research I did before getting one, the 980ti seems to be ~70% the performance of a 1080. So if you can find one for ~300 USD, it's still a heck of a deal, especially for 1080p/1440p.

EDIT: Especially if you overclock, or even get one of the factory overclocked versions. The 980ti FTW was for sale for 300 bucks for the holidays on EVGA's b-stock.

THIS. I did exhaustive research on reviews on here and other sites with the 980ti, 1070, and 1080 and made every comparison between models stock/overclocked in various games/scenarios etc. the "stock" 980ti is particularly overclockable and at pretty low clocks in relation to the 1000 series and factoring that in (since every card I'd overclock once I had it) especially at price I found it there was no choice. I WANTED to buy a new gen 1070 gigabyte mini because it's tiny and still oc'd fairly well and just so much less hassle moving around case and getting wires around it etc...was 399$ at the time. I went out of my way to search through multiple games/scenarios etc with the 1070 vs a 980ti with typical values when OC'd and honestly though I wanted that nice little mini card for space convenience and the newer tech/power consumption I looked at the numbers and I can safely say the 1070 can expect to lose pretty much every time when it and the ti are at max air clocks. At "stock" clocks the 980ti is 9% slower than 1070. You overclock them ti reverses that and then some.

If I didn't purchase this one I'd have gotten that b-stock that might be great card to monkey with and OC with good beefy circuitry and good EVGA cooler etc on it.

 

[EarthDog]

I looked at the numbers and I can safely say the 1070 can expect to lose pretty much every time when it and the ti are at max air clocks

If you overclock a 980Ti and a 1070 to their 'limits' (~1500 Mhz on the 980Ti and 2100 on the 1070), the 980Ti creeps closer, indeed, but it is still generally slower. I ran this test with a 1070 Gaming X and a 980Ti Extreme Gaming personally. Let me put it another way though...

1. 1070 vs 980Ti is 12% faster reference for reference
https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1070/24.html

2. The Gaming Z, is another 6% faster in its stock form (totaling 18%)
https://www.techpowerup.com/reviews/MSI/GTX_1070_Gaming_Z/26.html

3. That 1070 Gaming Z overclocked to 2114 MHz, adds more on top of that... in BF4, 11% on top of its stock clocks. If we went by that single title, the lead would be 39%. Since that isn't a good data set, I will cut in in half to 6% making the lead ~24% over a reference 980Ti. The 980Ti overclocks well, but it isn't going to catch a 1070 in most titles. The 980Ti will NOT "reverse that and then some". It will come very close though, buts its not taking a lead by several percent and pushing forward... no way. Been there, done this.
https://www.techpowerup.com/reviews/MSI/GTX_1070_Gaming_Z/29.html

 

[dalekdukesboy]

Simply taking that one game and randomly saying "I'll cut it in half to make 6% lead etc...yeah, that doesn't cut it. I'm talking USING good data sets and multiple sources but "reference" we already now and established and I'll note you didn't bother picking an OC'd factory ti to compare you just did that with 1070 to make your "point" such as it is. Also you throw a red herring out saying "I ran this test with a 1070 Gaming X and a 980Ti Extreme Gaming personally. Let me put it another way though..." Because then you include none of that data you claim you have so that is like me claiming I ran a test and the gtx 480 soundly beat a 1080 on my machine with my tests...and list my OC's and give no data to back it up.

 

[dalekdukesboy]

Also just looking up the one title Wiz uses to OC cards with I found the older driver tests with ti models vs newer drivers and 1070 on BF 3 same settings this is result....

I think it's fair to say that's a virtual tie....NOT withstanding about 1.5 years of driver updates the ti wasn't privy to at the time it was tested so if we want to just "pick" one title and play games at the least pick one where both are OC'd to the max (like this) and look at the raw numbers. Care to calculate the difference percentage wise there? I doubt you do.

 

[EarthDog]

I think its fair to say that your initial statement...

I can safely say the 1070 can expect to lose pretty much every time when it and the ti are at max air clocks....
....You overclock them ti reverses that and then some.

...was incorrect. Yes.

You can slice it any way you would like, but that statement was patently false. the 980Ti when fully overclocked on air, will NOT "reverse that (lead) and then some....every time". It will come closer, indeed. And I bet in a couple titles it takes the lead (didn't with our gaming suite at 1080p), but 'pretty much every time', and 'reverses it and then some' is being dramatic. You made the statements, its your job to back them up seeing how that is against what most believe to be true. Trust me, if I had a 980Ti still, I'd do it now.

A more fair statement can be that they trade punches. I would go along with that...but certainly not that it beats it every 'pretty much' every time..

 

[Ungari]

@Adored
Shows how similar the 980Ti is to the GTX 1080 at the same clock speeds, which also shows that the GTX 1070 is worse at the same clocks:

 

[dalekdukesboy]

You overclock them ti reverses that and then some.

At best, you can say "then some" is wrong in this instance I laid out, however going from your 12% to statistically 0% is pretty much eradicating the advantage and practically reversing the advantage no? Also, I could again point to driver maturation and that "reversal" may just happen and probably would considering I found a few tests with lower clocked ti's that with newer drivers got almost same number this one did at higher clocks. I will also point out I used your 1070 msi (which so happens to be the fastest OC achieved on site for 1070) and paired it with yes also the highest clocked 980ti but again I picked one game and one scenario and

"Since that isn't a good data set, I will cut in in half to 6% making the lead ~24% over a reference 980Ti"....Who's closer with hard data? This was patently absurd statement on it's face you even admitted you took the game that was terrible representative example and arbitrarily came up with this magic 6% number with no evidence to prove it meant anything to "prove" a point. I think your falsehoods, lack of data, overstatements and general perversion of data far exceeds mine particularly when you only can take one statement I said off the cuff and can't even blow that up because your data and thinking is so flawed.


You want to argue "generally" as you kept saying the 1070 as insinuated by this test with your generous 6% compensation somehow proves it's fair to say 24% is a representative number of difference between OC cards generally? I think not. You also huffed and puffed with your non existent data set you didn't even provide but simply "told" us you tested the two cards at x clocks and ti generally lost. Again, that statement is beyond false it isn't even verifiable to be false with zero data to accompany it.

 

[EarthDog]

I dont want to split hairs with you any more than I already have, honestly. I have no horse in this race except to make sure the people reading are armed with facts... or closer to reality.

Perhaps more carefully choose your words to mean what you say. Because 'reversing it' means that now the 980Ti has the same lead the 1070 did and then some means even more on top of that. Patently misleading at best... or cherry picking a result or two to make it true when the big picture tells a much different story. I know you wouldn't do that, so, just choose your words more wisely to mean more accurately the point you want to convey.



EDIT: Ahh ha!! Yup, some titles it does take the lead... others its still behind.. overall in this testing, at 2560x1440, it was still 3% behind on average.

1440P overclocked:
GTX 1070 is 3% faster than GTX 980Ti
GTX 1080 is 23% faster than GTX 980Ti

http://www.overclock.net/t/1601896/...ck-showdown-gtx-980ti-vs-gtx-1070-vs-gtx-1080

 

[dalekdukesboy]

I dont want to split hairs with you any more than I already have, honestly. I have no horse in this race except to make sure the people reading are armed with facts... or closer to reality.

Perhaps more carefully choose your words to mean what you say. Because 'reversing it' means that now the 980Ti has the same lead the 1070 did and then some means even more on top of that. Patently misleading at best... or cherry picking a result or two to make it true when the big picture tells a much different story. I know you wouldn't do that, so, just choose your words more wisely to mean more accurately the point you want to convey.



EDIT: Ahh ha!! Yup, some titles it does take the lead... others its still behind.. overall in this testing, at 2560x1440, it was still 3% behind on average.



http://www.overclock.net/t/1601896/...ck-showdown-gtx-980ti-vs-gtx-1070-vs-gtx-1080

I will reiterate one thing and one thing only...you better examine your own words and implications if you want to "split hairs" with me because you've not come close to proving me wrong...and if best you got is 3% on average. You essentially made my point for me. I appreciate the honesty to check further but indeed it does win "some titles" as you just said. I will say again, at best you can say I "may" be wrong it generally wins over 1070. You can't deny it does at times, and it is a razor thin advantage the 1070 has....that is not what you implied with the arbitrary 6% as well as implying 24% in one game is a fair metric to even use.

This was one of articles I read when I did my research on the card. However only one, and as pointed out by author cards had more in tank to go so that 3%....probably more like 0 and technically....3% is basically within test margin of error and repeatability etc. Again I appreciate you giving me numbers but at least admit you were pretty "off" by implying your 24% lead with the card etc and me implying it at least equaled the 1070 overclocked and yes beat it at times was accurate. So point is you are moving much closer to my statements in reference once you looked up numbers and much farther away where you were arguing from.

 

[revin]

Huh? So both cards clocked the same =about the same FPS on the game
So under clock the new card but why, so it just equals about the same tflop then well yes they should perform about the same.
WTH am I missing?????????????? rhetorical statement

 

[EarthDog]

I didn't make your point for you. My last link, the one you accept, shows overall, it still losing by 3%. Making your initial blanket statements patently false. I have quoted your posts and do not need any more clarity. I NEVER denied it could flip the table in a title or two... in fact I said, this in post 89:

And I bet in a couple titles it takes the lead

So anyway, that was a long way about things, but I think we can now safely say and agree the 980Ti is still a bit slower OVERALL when overclocked with the 1070, but can notch some victories in a couple of titles. A far cry from the initial statement.

 

[dalekdukesboy]

Huh? So both cards clocked the same =about the same FPS on the game
So under clock the new card but why, so it just equals about the same tflop then well yes they should perform about the same.
WTH am I missing?????????????? rhetorical statement

It's just a way of showing mhz for mhz one is faster. Admittedly that isn't what I was going for but only max clocks for both, which puts us at a virtual tie and trading blows. My point is the 1070 is NOT really statistically faster than a 980ti when both are clocked to max. I still would like a test NOW not from June and see where they'd end up AND with 1500 mhz on the ti and 2100 on the 1070. Regardless at best I think it's fair to say the 980ti =1070 when overclocked. NOT that the ti is faster but also my point definitely NOT that the 1070 is faster and appreciably so like...say 24%.

 

[EarthDog]

It's just a way of showing mhz for mhz one is faster. Admittedly that isn't what I was going for but only max clocks for both, which puts us at a virtual tie and trading blows. My point is the 1070 is NOT really statistically faster than a 980ti when both are clocked to max. I still would like a test NOW not from June and see where they'd end up AND with 1500 mhz on the ti and 2100 on the 1070. Regardless at best I think it's fair to say the 980ti =1070 when overclocked. NOT that the ti is faster but also my point definitely NOT that the 1070 is faster and appreciably so like...say 24%.

They may end up further apart if tested now. Haven't people been complaining about the lack of improvements in older generations???

I never said it was 24% faster than the 980Ti overclocked... I was trying to show you how much improvement the 1070 can have over its own reference clocks. You didn't like that OC data set to get to 24%, understandably since its one title - but its the best I had, so I found another that proves its a lot closer than your initial statement which had the 980Ti being the clear winner when overclocked. Cheers.

 

[dalekdukesboy]

So anyway, that was a long way about things, but I think we can now safely say and agree the 980Ti is still a bit slower OVERALL when overclocked with the 1070, but can notch some victories in a couple of titles. A far cry from the initial statement. [/QUOTE]

You moved farther to my point than yours, so last I checked that means I was more accurate.

Let's review and see who was closest shall we?

980Ti doesn't even match the 1070, forget the 1080...

Hmm...well throw that out

True, 980ti doesn't even come close to 1080. About 30% slower from what I've seen. But it does match the 1070, nearly identically. And can be had for 100 bucks cheaper.

I think we have a winner.

Fair enough... We could argue aaaaall day long, and post benchmark after benchmark. Ive found several where the 980ti even beat the 1070. But the truth is, at 1080p and 1440p, real-world performance difference that you can truly see, is going to be negligible.

and again...

THIS. I did exhaustive research on reviews on here and other sites with the 980ti, 1070, and 1080 and made every comparison between models stock/overclocked in various games/scenarios etc. the "stock" 980ti is particularly overclockable and at pretty low clocks in relation to the 1000 series and factoring that in (since every card I'd overclock once I had it) especially at price I found it there was no choice. I WANTED to buy a new gen 1070 gigabyte mini because it's tiny and still oc'd fairly well and just so much less hassle moving around case and getting wires around it etc...was 399$ at the time. I went out of my way to search through multiple games/scenarios etc with the 1070 vs a 980ti with typical values when OC'd and honestly though I wanted that nice little mini card for space convenience and the newer tech/power consumption I looked at the numbers and I can safely say the 1070 can expect to lose pretty much every time when it and the ti are at max air clocks. At "stock" clocks the 980ti is 9% slower than 1070. You overclock them ti reverses that and then some.

Of that whole paragraph, at best "then some" is all you would throw out. I think going from double digits say....24% to 3% is pretty much reversal of fortunes no?
Particularly since that is still only ONE test with chosen games etc.

If you overclock a 980Ti and a 1070 to their 'limits' (~1500 Mhz on the 980Ti and 2100 on the 1070), the 980Ti creeps closer, indeed, but it is still generally slower. I ran this test with a 1070 Gaming X and a 980Ti Extreme Gaming personally. Let me put it another way though...

1. 1070 vs 980Ti is 12% faster reference for reference
https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1070/24.html

2. The Gaming Z, is another 6% faster in its stock form (totaling 18%)
https://www.techpowerup.com/reviews/MSI/GTX_1070_Gaming_Z/26.html

3. That 1070 Gaming Z overclocked to 2114 MHz, adds more on top of that... in BF4, 11% on top of its stock clocks. If we went by that single title, the lead would be 39%. Since that isn't a good data set, I will cut in in half to 6% making the lead ~24% over a reference 980Ti. The 980Ti overclocks well, but it isn't going to catch a 1070 in most titles. The 980Ti will NOT "reverse that and then some". It will come very close though, buts its not taking a lead by several percent and pushing forward... no way. Been there, done this.
https://www.techpowerup.com/reviews/MSI/GTX_1070_Gaming_Z/29.html

So....you think this is close to "winning" the argument? I think not. Flawed logic, lacking data and implying a 24% lead is any way typical for both at max clocks...you think you proved your point finding an average of 3% when the closest you imply it gets is 12% at stock? You said very close...yeah that is understated and yes it does push ahead in some scenarios which you said "no way". I think 3% is closer than very close, again that is within test margins so properly stated it would be an essential tie as you did properly restate later that they trade blows almost equally. "creeps closer" is a vast understatement as well, that implies a sound distance between the two cards at OC...3% is up the 1070's ass.

 

[t77snapshot]

Hey guys, what is your opinion on the 780ti for games like bf1 nowadays?

Off topic or not this is good stuff, I am enjoying all the input. Still haven't gotten a card yet, looking to only spend $200 or less though so I can't go crazy.

Personally I wouldn't buy a used card regardless the condition of it, but hey that's just me.
I'd say get a new GTX1060 6GB non reference.

I've been pc gaming for 15 years and never once bought a brand new graphics card lol. In my option hardware can go a long way if taken care of.

 

[ASOT]

Great card for 1080p still nowdays a good performer,if not exceed those 3gb