Battlefield 1 Benchmarks (Gamegpu & the rest)

[RussianSensation]

Thanks RussianSensation very interesting!

Is there any advantage a GTX 1070 has then?

Yes, many. GTX1070 is better than Fury in every way, except price/performance. I'll make it simpler for you. If you don't mind reselling your GPU in 18 months and buying something faster, and BF1 is your main game for the next 6-9 months, get the Fury. If you want to keep your card for 3 years, play a wide variety of games, get the 1070. One caveat is if you have a 1080p 60Hz monitor and an i3/i5. In that case, I cannot recommend a $380 1070 over a $225 Fury. The extra performance for such a disproportionate price increase will be largely wasted for most of the useful life of the 1070. If you are gaming at 1080p 60Hz, get a $225 Fury and buy a new videocard in 2-2.5 years instead.

 

[Bacon1]

Sketchy that NV's GPUs lose so much performance in the DX12 codepath.

Wait for Nvidia drivers

Thanks RussianSensation very interesting!

Is there any advantage a GTX 1070 has then?

I found a eVGA Video Card GTX 1070 ACX 3.0 8GB DDR5 256Bit PCI Express Retail for $379.99 and free shipping.

I wouldn't buy it over those super cheap Fury at all. Heck for the price difference you could buy a brand new freesync monitor to pair with it.

 

[pcslookout]

Yes, many. GTX1070 is better than Fury in every way, except price/performance. I'll make it simpler for you. If you don't mind reselling your GPU in 18 months and buying something faster, get the Fury. If you want to keep your card for 3 years, get the 1070.

Thanks but how is the GTX1070 better than the Fury in every way, except price/performance ? Is the Fury faster than the GTX1070 ? If so I would be surprised. I thought they were very close in performance.

 

[RussianSensation]

Thanks but how is the GTX1070 better than the Fury in every way, except price/performance ? Is the Fury faster than the GTX1070 ? If so I would be surprised. I thought they were very close in performance.

price/performance = Price-to-performance. Let's not derail this thread into Fury vs. 1070. Go into your main thread to talk about it. Just look at the actual FPS in the games you play or intend to play at your resolution to see if Fury is fast enough:
https://www.techpowerup.com/reviews/MSI/GTX_1070_Gaming_Z/6.html

For BF1, Fury and 1070 are going to be very similar in performance.

//// back to BF1 discussion.

 

[Arachnotronic]

Wait for Nvidia drivers

Just going to use the DX11 codepath.

 

[pcslookout]

price/performance = Price-to-performance. Let's not derail this thread into Fury vs. 1070. Go into your main thread to talk about it. Just look at the actual FPS in the games you play or intend to play at your resolution to see if Fury is fast enough:
https://www.techpowerup.com/reviews/MSI/GTX_1070_Gaming_Z/6.html

For BF1, Fury and 1070 are going to be very similar in performance.

//// back to BF1 discussion.

Ok thanks. Sorry guys.

 

[pcslookout]

Ok i made a topic just for my question
RussianSensation

http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=thread...apphire-fury-be-a-worthwhile-upgrade.2488975/

 

[Carfax83]

Since the game hasn't been released, it's too early to determine final performance conclusions. Actually, it will probably take months after release before final performance optimizations are attained, if BF4 is anything to go by. That said, I'm continually baffled by developers' inability to beat NVidia's DX11 driver optimizations. NVidia's DX11 driver mimics DX12 somewhat in what it does, so the bar is significantly higher than AMD's atrocious CPU overhead in their DX11 driver..

But I would have thought that DICE being amongst the highest caliber of software engineers would have overcome the hurdle, especially at this stage.. Perhaps my faith in DICE was misplaced?

 

[Arachnotronic]

Since the game hasn't been released, it's too early to determine final performance conclusions. Actually, it will probably take months after release before final performance optimizations are attained, if BF4 is anything to go by. That said, I'm continually baffled by developers' inability to beat NVidia's DX11 driver optimizations. NVidia's DX11 driver mimics DX12 somewhat in what it does, so the bar is significantly higher than AMD's atrocious CPU overhead in their DX11 driver..

But I would have thought that DICE being amongst the highest caliber of software engineers would have overcome the hurdle, especially at this stage.. Perhaps my faith in DICE was misplaced?

No, DICE is quite good, among the best in the business. But it's really, really tough to beat NVIDIA's driver teams.

 

[crisium]

Interesting to see the perfect Fury X vs 290X TFLOP scaling.

Is this an end to the "Fiji front end bottleneck" idea and whenever we do not see this scaling it's simply fail drivers? Or are the 290X drivers fail here? I think it's the latter.

I think Hawaii isn't performing here. 290X = 470 isn't quite right. Also 290 is only 4-8% faster than the 380X at 1440p DX11-12. That's called a red flag, folks.

Odd to see Hawaii suffer when it often does the best seeing as its closest architecturally to whats in the consoles.

 

[Maverick177]

Very nice performance for all the cards. This is what happens when a game is properly coded.

 

[Mercennarius]

Interesting to see the perfect Fury X vs 290X TFLOP scaling.

Is this an end to the "Fiji front end bottleneck" idea and whenever we do not see this scaling it's simply fail drivers? Or are the 290X drivers fail here? I think it's the latter.

I think Hawaii isn't performing here. 290X = 470 isn't quite right. Also 290 is only 4-8% faster than the 380X at 1440p DX11-12. That's called a red flag, folks.

Odd to see Hawaii suffer when it often does the best seeing as its closest architecturally to whats in the consoles.

Agreed. But I wonder are they using a Reference 290(x) in their testing? There's a pretty big performance increase going to a non-reference overclocked 290/390 vs the reference design.

 

[Carfax83]

No, DICE is quite good, among the best in the business. But it's really, really tough to beat NVIDIA's driver teams.

I guess we'll see when the final game ships in a week with release day drivers..

 

[nurturedhate]

I guess we'll see when the final game ships in a week with release day drivers..

For all intents and purposes the game released today for EA Access members.

 

[Sweepr]

Hmm Fury NANO 175W TDP 28nm = GTX 1070 150W TDP 16nm at 1440p DX-12

Insane perf/watt for 28nm product.

GTX 1070 DX11 performance is the same as Fury X in DX12 while using 130W less power.

 

[Bacon1]

GTX 1070 DX11 performance is the same as Fury X in DX12 while using 130W less power.

Try using an AIB 1070:

Not to mention that PEAK isn't representive of normal usage:

 

[Sweepr]

So still a 85W+ difference picking the most power hungry GTX 1070s and comparing it to reference Fury Xs, got it. Not impressed.

Edit - Some custom cards draw less power than FE:

 

[Maverick177]

So still a 85W+ difference picking the most power hungry GTX 1070s and comparing it to reference Fury Xs, got it. Not impressed.

Well when you compared a last gen product with the newer gen, this tends to happen. For me, I'm not impressed about Pascal's performance , Fiji on the other hand....

 

[Snarf Snarf]

So still a 85W+ difference picking the most power hungry GTX 1070s and comparing it to reference Fury Xs, got it. Not impressed.

The question was about Fury vanilla air, not Fury X. Also there is a whole thread about this discussion started by pcslookout if you want to continue it over there.
http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=threads/best-video-card-for-battlefield-1.2488752/

 

[nurturedhate]

The question was about Fury vanilla air, not Fury X. Also there is a whole thread about this discussion started by pcslookout if you want to continue it over there.
http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=threads/best-video-card-for-battlefield-1.2488752/

Don't send him over there. That thread is still relatively peaceful atm.

 

[sontin]

Sweclockers has benchmarks, too: http://www.sweclockers.com/test/22807-snabbtest-battlefield-1-i-directx-11-och-directx-12

Another Gaming Evolved game with a broken DX12 path.

 

[Snarf Snarf]

Sweclockers has benchmarks, too: http://www.sweclockers.com/test/22807-snabbtest-battlefield-1-i-directx-11-och-directx-12

Another Gaming Evolved game with a broken DX12 path.

Performance numbers look pretty good actually, Nvidia is probably going to gain a good 5-10% once there's a game ready driver or a patch.

Would hardly call this a broken DX12 path.

Edit. Woof that i5 looks like its getting hammered. The other benches were done on i7's and the difference shows

 

[IllogicalGlory]

My god, an extra 85W? You may as well be turning your computer into a furnace! Better spend that extra $160. In truth, the difference is only 55W anyway even if it's the Nitro model, which is nothing.

Shame that FUD about BF1 requiring 8GB didn't come true.

 

[sontin]

Performance numbers look pretty good actually, Nvidia is probably going to gain a good 5-10% once there's a game ready driver or a patch.

Would hardly call this a broken DX12 path.

You need to read more about DX12 at this point. It is useless in the multiplayer right now.
Lets see what will happen next week.

 

[nurturedhate]

So still a 85W+ difference picking the most power hungry GTX 1070s and comparing it to reference Fury Xs, got it. Not impressed.

Great way to move goalposts. We all see it, nice try though. The conversation is about reg fury. The difference is 25-55 watts. Go change a lightbulb and save twice that. That flawed argument has been beaten into the ground, let alone being a complete non issue during Fermi.

Secondly, what's with your "reference" Fury X comment? Was that some form of failed insult? Like "Hey look at this terrible AIO on a reference card, look how quite it is, it's terrible, I should know". Come on. The issue here is that your usual rhetoric of AMD bad, Nvidia is God doesn't work here. Both companies do well in this game. You are left with grasping at straws.