[Update] DX 11 vs DX12 benchmarks addded - Gigabyte reveals Battlefield 1 Benchmarks! (General BF1 benchmark thread)

[AlTech]

UPDATE 1:

 

WCCFTECH has released an article similar to this about the DX 11 and DX 12 performance....

 

AND AMD is making the gains while Nvidia is losing...

 

 

Original Article:

Gigabyte Technologies, the company behind Gigabyte AIB GPUs, has revealed the benchmarks for Battlefield 1.

 

Without further a do, here are the benchmarks........

 

Specs used by Gigabyte

Quote

Motherboard: GIGABYTE GA-X99-SLI

• CPU: Intel Core i7-5960X
• Memory: 16 GB DDR4 Memory
• Operating System: Windows 10

 

 

 

Quote

At the 1440p resolution, things start to heat up. On the low-endside, only the GTX 960 can hit the 1440p 30 standard, the RX 460 drops out of the race. All other cards fare well enough. Once we upgrade to the 1440p 60 standard however, only the GTX 1070 and the GTX 1080 can keep up at 70 and 91 fps respectively. The RX 470 and the RX 480 from the AMD side and the GTX 960, GTX 970 and the GTX 1060 from Nvidia’s side can no longer keep pace (although both the RX 480 and the GTX 1060 are close enough at 50 and 53 fps respectively).

At the 4K resolution, the graphics cards were only able to achieve the 4K 30 standard. The 4K 60 territory remains uncharted for a single card. The GTX 1070 scores an average of 34 fps in the 4K benchmark while its bigger brother, the GTX 1080, is able to score a slightly more appetizing 48 fps – which is pretty close to a full 60 fps. It is clear that we are just 1 generation away from achieving true 4k 60 performance in the high end – and maybe that’s where the GTX 1080 Ti will step in. At this point, however, this is something that hasn’t happened. Of course, Battlefield 1 is a very demanding title, both the GTX 1070 and 1080 should be able to hit the 4k 60 mark in less demanding titles – but what’s the fun in that.

 

Also, some things to note

Quote

This benchmark does not include the Fury series from AMD which, while old, currently fills the high-end spectrum for Radeon GPUs. It goes without saying that they will perform better than the Polaris lineup and will be able to achieve performance somewhere between the GTX 1070 and 1080. We will be doing a follow-up post on Battlefield 1 benchmarks which will contain all currently available GPUs, not just FinFET based ones.

 

 

 

What do you think? This seems ok. But I would have liked to seen how older cards performed.

 

http://wccftech.com/battlefield-1-benchmarks-nvidia-takes-lead-gtx-1070-gtx-1080/

[AlTech]

1 minute ago, wrathoftheturkey said:

Yes, we know who Gigabyte is, m8. They make a lot of other stuff too.

I always give a short description of who they are in case somebody doesn't know who I'm talking about.

[VerticalDiscussions]

Well, i wouldnt choose any other GPU retailer from Nvidia that wasnt EVGA cooler designs or reference card + EKWB block. For AMD, Sapphire all the way aswell.

 

Eh, this Battlefield 1 game seems to be well made then eh? Not a terrible port it would seem.

[Mr.Meerkat]

So the Rx480 beats the 1060 at med and high settings but the 1060 beats the 480 at low and ultra settings...wut?  

[AlTech]

Just now, Mr.Meerkat said:

So the Rx480 beats the 1060 at med and high settings but the 1060 beats the 480 at low and ultra settings...wut?  

It narrowly beats the 480 at low and ultra. They perform almost the same.......

[Mr.Meerkat]

2 minutes ago, AluminiumTech said:

It narrowly beats the 480 at low and ultra. They perform almost the same.......

True true  

[Guest]

4 minutes ago, Mr.Meerkat said:

So the Rx480 beats the 1060 at med and high settings but the 1060 beats the 480 at low and ultra settings...wut?  

[AresKrieger]

One thing to note the 480 has very low dips (by comparison) and a higher max fps on ultra meaning it's is quite unstable fps wise possibly explaining the odd results for differing settings

 

Also why didn't they use a g1 gaming 970? It performs much better than the reference and they did so for other cards, doesn't that seem odd @AluminiumTech

[ivan134]

18 minutes ago, Mr.Meerkat said:

So the Rx480 beats the 1060 at med and high settings but the 1060 beats the 480 at low and ultra settings...wut?  

While the results don't make any sense, I would still ignore these results, especially for 480 because their G1 480 is garbage. Worse components than the reference 480 and reaches 95c even on an open test bench 

[Zyndo]

1 minute ago, ivan134 said:

While the results don't make any sense, I would still ignore these results, especially for 480 because their G1 480 is garbage. Worse components than the reference 480 and reaches 95c even on an open test bench 

are you referring to the JayztwoCents video? coulda just been a bad card. are there others having this issue as well?

[ivan134]

1 minute ago, Zyndo said:

are you referring to the JayztwoCents video? coulda just been a bad card. are there others having this issue as well?

Yes, that plus user feedback on reddit and other tech forums. Watch buildzoid's breakdown of the g1 480's PCB. That card is just garbage.

[Avcn42]

VRAM usage?

[BuckGup]

So no 980ti? I am guessing it would make the 1070 and 1080 look bad. But whatever they included the 970

[Zeeee]

well..... im running it ultra EVERYTHING MAX even aa on taa... at 1440p and i get 90-100fps  others gtx 1070s need to get a check up, maybe its gigabyte that is terrible

[Kinda Bottlenecked]

22 minutes ago, SamStrecker said:

So no 980ti? I am guessing it would make the 1070 and 1080 look bad. But whatever they included the 970

 

I'd say it would slot right in between the 1070 and 1080. 

[Energycore]

3 hours ago, AluminiumTech said:

I always give a short description of who they are in case somebody doesn't know who I'm talking about.

It's good journalism to do it so I say do it

[Energycore]

13 minutes ago, Pohernori said:

 

I'd say it would slot right in between the 1070 and 1080. 

It will probably be below the 1070, but the Fury X is one that might perform great or not.

[Kinda Bottlenecked]

1 minute ago, Energycore said:

It will probably be below the 1070, but the Fury X is one that might perform great or not.

 

Ah right, i forgot peeps only bechmark the 980 tis at 1Ghz. 

[AlTech]

3 hours ago, AresKrieger said:

One thing to note the 480 has very low dips (by comparison) and a higher max fps on ultra meaning it's is quite unstable fps wise possibly explaining the odd results for differing settings

 

Also why didn't they use a g1 gaming 970? It performs much better than the reference and they did so for other cards, doesn't that seem odd @AluminiumTech

Yeah. It seems odd.

 

Almost like they didn't want the 970 to be portrayed very well.

 

55 minutes ago, SamStrecker said:

So no 980ti? I am guessing it would make the 1070 and 1080 look bad. But whatever they included the 970

Yeah. These benchmarks are biased towards the newer cards. If they're excluding older cards, it's to make people think they need to buy newer cards.

 

 

[BuckGup]

49 minutes ago, Pohernori said:

 

I'd say it would slot right in between the 1070 and 1080. 

Yeah making their new products look bad

[Techicolors]

pretty strange they didn't bother testing fury cards, but ironically they think it's between the 1070 and 1080. 

[YongKang]

Tweaktown benchmarks coming soon this weekend. Let's wait guys.

[Bouzoo]

42 minutes ago, YongKang said:

Tweaktown benchmarks coming soon this weekend. Let's wait guys.

You mean same guys who said Kaby Lake is 40% faster than Skylake?

I like this WCCF journalism. They say no card can vice 4K/60, yet they don't mention Titan X. 

I'll wait for Guru3D and few other sites that I trust. 

Overall, the game runs pretty good overall. Were the drivers even released on any side? 

And drop the AA on 4K.

[ThinkWithPortals]

12 hours ago, AluminiumTech said:

 

Those RX 480 results are noice, should be able to hit 60fps stable on High on my 2560x1080 display. Perhaps I'll sign up for Origin around Christmas, try out BF1 and Overwatch.

[jaggysnake57]

you have to givie it to dice....this game looks fucking amazing for how much power it needs to run it