• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

(gamegpu) Battlefield 1 Benchmarks

F

Final8ty

Gawd
Joined
Jun 13, 2006
Messages
1,001
QjIiDv.png

SaCsaj.png


48u3wL.png

cEdaI4.png


WvCOdp.png

yCtAib.png


Jdzy6K.png

IYvivq.png
Battlefield 1 тест GPU | Action / FPS / TPS | Тест GPU
 
Minimalistic difference between DX11 and DX12 for both CPU and GPU. And again no mGPU support in DX12. That was disappointing from DICE.
 
I know this is the videocard subforums, but when in the hell is Intel or AMD going to release a faster CPU? I hope Kaby Lake doesn't sink and Zen isn't pretend. I think I might end up upgrading for the hell of it, because it seems like it won't be for more performance.

IYvivq.png



luke51087 said:
anyone else surprised at the 480vs 1060 6gb?
Click to expand...

Well it might be drivers or the near granite assumption that Nvidia's hardware doesn't like DX12. I really think it's drivers since AMD doesn't have a CrossfireX driver out yet. Maybe Nvidia just hasn't released their Battlefield 1 performance driver also.
 
cageymaru said:
I know this is the videocard subforums, but when in the hell is Intel or AMD going to release a faster CPU? I hope Kaby Lake doesn't sink and Zen isn't pretend. I think I might end up upgrading for the hell of it, because it seems like it won't be for more performance.

IYvivq.png





Well it might be drivers or the near granite assumption that Nvidia's hardware doesn't like DX12. I really think it's drivers since AMD doesn't have a CrossfireX driver out yet. Maybe Nvidia just hasn't released their Battlefield 1 performance driver also.
Click to expand...

On that note, I don't think AMD has released their BF1 driver yet either.
 
It is still pretty interesting on raw perf, the 480 has an advantage in dx11 at low res. Who'd have thought that?
 
the beta ran great for me, i'm ready to go for the full release! excited for this game.
 
Shintai said:
Minimalistic difference between DX11 and DX12 for both CPU and GPU. And again no mGPU support in DX12. That was disappointing from DICE.
Click to expand...
But...but..you need DX12, its the only way you should play next gen games and the only OS the supports it is Windows 10. Which is designed for next gen games because of DX12. Did i mention Windows 10 with DX12?
 
cageymaru said:
I know this is the videocard subforums, but when in the hell is Intel or AMD going to release a faster CPU? I hope Kaby Lake doesn't sink and Zen isn't pretend. I think I might end up upgrading for the hell of it, because it seems like it won't be for more performance.
Click to expand...

I love the fact my 3 year old 4770k is still a top gaming CPU. I mean really, not even a single FPS increase going to a 6700? Maybe in another 3 years I'll have a reason to upgrade!
 
AMD's advantage here is a lot less than previous DICE games. They were up some 20% in Battlefront.
I suppose the AMD community will complain about being gimped?

Surprisingly, DX12 gains almost non-existant (within margin of error).

This will get chalked up as a win for AMD despite how well Nvidia did this time around. After a long string of losses they needed this one, good thing DICE is on team red.
At least the game runs fine.

Interested in other sources as GameGPU cares more about being first than accurate.
 
Last edited:
TaintedSquirrel said:
AMD's advantage here is a lot less than previous DICE games. They were up some 20% in Battlefront.
I suppose the AMD community will complain about being gimped?

Surprisingly, DX12 gains almost non-existant (within margin of error).

This will get chalked up as a win for AMD despite how well Nvidia did this time around. After a long string of losses they needed this one, good thing DICE is on team red.
At least the game runs fine.

Interested in other sources as GameGPU cares more about being first than accurate.
Click to expand...

I'm sorta confused. You want some accurate benchmarks for a game that neither Nvidia nor AMD have released drivers for, and that isn't even out for general release yet. (Aside from the various tiers of release (i.e Orgin Access, Deluxe Edition access) being an absolute cash grab, while I'm not thinking there will be performance problems like in BF3, I do think this will allow DICE time to fix whatever problems start rising up before the general masses start playing)
 
Last edited:
I like how sometimes the graphs show a higher number with a lower bar...someone needs to learn how to make graphs, or it's a concerted effort to make the AMD offerings look weak against NVidia cards.

Conspiracy theories aside, it usually does show the AMD parts in a negative light VS NVidia.

TaintedSquirrel said:
Surprisingly, DX12 gains almost non-existant (within margin of error).
Click to expand...

Is it really that surprising that engines developed before DX12 don't show any DX12 gains? Come on now.
 
buzzbomb said:
Is it really that surprising that engines developed before DX12 don't show any DX12 gains? Come on now.
Click to expand...
From the company that designed Mantle, sure. See: Battlefield 4.
If there's one game dev who is perfectly situated to showcase the benefits of DX12, it's DICE.
 
DX12 has elements of Mantle in it, but it isn't Mantle. The current iteration of Frostbite is approaching 2 years old - not that that's a bad thing, if it works it works - but devs aren't keen on writing new engines when they don't absolutely need to do so.
 
buzzbomb said:
DX12 has elements of Mantle in it, but it isn't Mantle. The current iteration of Frostbite is approaching 2 years old - not that that's a bad thing, if it works it works - but devs aren't keen on writing new engines when they don't absolutely need to do so.
Click to expand...
I don't follow. Ashes started off as a Mantle game and ended up being the poster boy for DX12 last year.
Are we, perhaps, digging deep to make excuses on AMD's behalf? I mean, it seems that way.
 
The thing I am shocked no one talks about. The Fury X being only 6fps slower than the 1080 GTX @ 1440p....with no proper driver release.

I am sorry but if that doesn't show how good AMD's cards are at DX12 is I don't know what will.
 
NukeDukem said:
I love the fact my 3 year old 4770k is still a top gaming CPU. I mean really, not even a single FPS increase going to a 6700? Maybe in another 3 years I'll have a reason to upgrade!
Click to expand...

That is also without even being overclocked.

Thats why I am still on 1366. Right now the only people who truly need more FPS are the 120/144hz players. Thats when it comes to the CPU.
 
NukeDukem said:
I love the fact my 3 year old 4770k is still a top gaming CPU. I mean really, not even a single FPS increase going to a 6700? Maybe in another 3 years I'll have a reason to upgrade!
Click to expand...
because they're GPU/API bottlenecked in the DX12 CPU performance graph... when the bottleneck is removed in DX11 the 6700 (which is a good bit slower than a 6700K btw) gains a significant lead over the 4770K. same deal with the 5960X, it should be above everything else but for some reason they didn't use 1080 SLI.
 
Brackle said:
The thing I am shocked no one talks about. The Fury X being only 6fps slower than the 1080 GTX @ 1440p....with no proper driver release.

I am sorry but if that doesn't show how good AMD's cards are at DX12 is I don't know what will.
Click to expand...
The 1080 also loses 10% performance switching from DX11 to DX12 at 1440p, something that shouldn't be happening. We've seen it gain performance in others.
 
TaintedSquirrel said:
The 1080 also loses 10% performance switching from DX11 to DX12 at 1440p, something that shouldn't be happening. We've seen it gain performance in others.
Click to expand...

Possibly software async compute vs hardware async compute?
 
Last edited:
Shintai said:
Snabbtest: Battlefield 1 i DirectX 11 och DirectX 12 - Test

In their test, all cards gets worse minimums in DX12. For AMD users its pretty much unplayable minimums.
bf11.jpg

bf12.jpg

bf13.png
Click to expand...
Looks like the days of the I5's are coming to a close. Those minimums are much lower then with the 5960x where AMD cards have higher minimums with DX 12. With limited threaded CPU's DX 12 does not do as well which is expected. Anyways looks like AMD does better with DX 12 then DX 11 while Nvidia does worst with DX 12 over DX 11 looking at the first benchmarks in the thread. Plus seeing some game specific addressed drivers may review more.

So if you are an AMD user go with DX 12 for the better experience unless you are on a limited threaded cpu. Nvidia stick with the older API of DX 11. Fiji does compete well with the 1070 Pascal in this game. For those that this game will be their primary game for the next year or so (there will be plenty doing this) - getting a cheap Fury or FuryX maybe a good buy.
 
Final8ty said:
It does look like amount of software async compute being offloaded to the CPU in DX12 in this game which is slowing down the 1080 in this game.
Click to expand...

Why would it be on the CPU? Did you just make up random FUD? :)

AMD does tessellation on the CPU, right? Or is that equally as wrong.
 
Last edited:
noko said:
Looks like the days of the I5's are coming to a close. Those minimums are much lower then with the 5960x where AMD cards have higher minimums with DX 12. With limited threaded CPU's DX 12 does not do as well which is expected. Anyways looks like AMD does better with DX 12 then DX 11 while Nvidia does worst with DX 12 over DX 11 looking at the first benchmarks in the thread. Plus seeing some game specific addressed drivers may review more.

So if you are an AMD user go with DX 12 for the better experience unless you are on a limited threaded cpu. Nvidia stick with the older API of DX 11. Fiji does compete well with the 1070 Pascal in this game. For those that this game will be their primary game for the next year or so (there will be plenty doing this) - getting a cheap Fury or FuryX maybe a good buy.
Click to expand...

I disagree, look at the i3 6100 placement. Above 6 and 8 cores. Or the i5 6600. The problem is rather that older architectures are going out of focus for developers. Just as they do on graphics. Its not because one age better or worse as such from a hardware standpoint or development catches up.

I wouldn't say AMD does better in DX12, look at sweclockers again. Those minimums! 16-21 FPS minimums depending on AMD card. While they do 35-41 in DX11. Its an easy choice isn't it?

Had it been most other companies it would be rather easy to dismiss. But this is DX12 from DICE that hyped it to the sky. And in the end of the day, you want to play on DX11 for both AMD and Nvidia for better performance.

And hopefully you dont buy
 
Last edited:
i5 6600 is pulling 90~100fps min, I'd say it's fine.
Although I agree 4c/4t chips are showing more trouble signs these days. The 2700K pulling ahead of a brand new i5... Yeah.
 
i3 6100 looks like a really nice cpu for budget or htpc configs. Same performance level as 2500k with its low TDP for mere $100 is great. Pair it up with upcoming 1050ti and you've got a 1080p console killer.
 
Raendor said:
i3 6100 looks like a really nice cpu for budget or htpc configs. Same performance level as 2500k with its low TDP for mere $100 is great. Pair it up with upcoming 1050ti and you've got a 1080p console killer.
Click to expand...
b1_intel.png


b1_amd.png



No, I wouldn't recommend any 4 thread max CPU for this. Even AMD 6 thread is dragging pretty hard.
 
jbc029 said:
No, I wouldn't recommend any 4 thread max CPU for this. Even AMD 6 thread is dragging pretty hard.
Click to expand...

Don't forget that all these cpus are tested with 1080. Of course at 1080p with GTX 1080 a cpu like i3 will be pegged and become a bottleneck. However if you pair it up with a card like 1050ti which level of performance those i3 were designed for in the first place, you won't see such levels of utilization.
 
Raendor said:
Don't forget that all these cpus are tested with 1080. Of course at 1080p with GTX 1080 a cpu like i3 will be pegged and become a bottleneck. However if you pair it up with a card like 1050ti which level of performance those i3 were designed for in the first place, you won't see such levels of utilization.
Click to expand...

A 380X (which is measurably stronger than a 1050ti) can barely hold onto 60 fps avg @ 1080p with absolutely no CPU limitation. Somehow I think substantially constraining the strength of the CPU won't improve that situation.
 
If you plan to play DX12 I would not recommend an I5/4core CPU. You will need HT for sure.

Also remember these are not done in MP. When you play on a 64 man server the CPU is pegged hard. So that as well will kill a 4core cpu.

DX12 is the way to go. Those minimums are fucking amazing!

And the Nano beating a 980ti in DX12....DAAAAMN
 
jbc029 said:
A 380X (which is measurably stronger than a 1050ti) can barely hold onto 60 fps avg @ 1080p with absolutely no CPU limitation. Somehow I think substantially constraining the strength of the CPU won't improve that situation.
Click to expand...
380X is stronger than a 1050Ti, let alone measurably? Source?
 
If its 20% faster than a GTX960 across the board, its also faster than 380/380X in average games.
 
lolfail9001 said:
Well, just look at how it is at least 30% to 50% faster than rx460 in same benchmarks you mention. If that lead holds up in real life, AMD may close shop with P11.

Is 380x actually even faster than rx460 than that in reality? No? QED.
Click to expand...
Yes. A 380 (non X) is substantially faster than a 460, the 380X even further ahead. The non X 380 is actually right around where the 1050ti seems to be. That performance level isnt "console killer". It's comparable.
 
jbc029 said:
Yes. A 380 (non X) is substantially faster than a 460, the 380X even further ahead. The non X 380 is actually right around where the 1050ti seems to be. That performance level isnt "console killer". It's comparable.
Click to expand...
Sure, another point of view to discard your premise of synthetics (especially those) being relevant: 960 is generally equal with 380 (faster on slow CPUs, slower on overclocked 6700ks). Since 1050Ti is notably faster it means that it is probably faster than 380X. What now?
 
lolfail9001 said:
Sure, another point of view to discard your premise of synthetics (especially those) being relevant: 960 is generally equal with 380 (faster on slow CPUs, slower on overclocked 6700ks). Since 1050Ti is notably faster it means that it is probably faster than 380X. What now?
Click to expand...
Source?
 
You must log in or register to reply here.
Back
Top