Looks to me as though 1600p is ok for 2gb cards though you may run in to trouble if you add 4xmsaa
To be honest though, i'm not 100% convinced. I want to see more benchmarks first.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
***THE BF4 BENCHMARKS THREAD***
- Thread starter pgi947
- Start date
Looks to me as though 1600p is ok for 2gb cards though you may run in to trouble if you add 4xmsaa
To be honest though, i'm not 100% convinced. I want to see more benchmarks first.
Soldato
I'd strongly recommend waiting for retail...least then you can have a look at the new AMD line up too. Saying that I doubt much will change, BF3 was always memory hungry which is why all this disbelief at the requirements for 4 leave me perplexed.
Caporegime
Pretty much yes.
I believe x4AA at 1080p should be possible with 2gb providing you have the grunt. The real test will be a long gameplay session. Its no good playing for 3 minutes and assuming it works ok. A quick look at the gamegpu and swe clockers benchmarks will tell you that.
I believe that at the moment we don't have FULL ULTRA settings. I think some things have been left out. I would expect vram usage to increase on the final game. Same happened with bf3 last year. You could select Ultra, but you only got high settings.
This year if you select Ultra preset you get mainly ultra with a few high settings. I've tried changing the settings to all ULTRA and get no fps difference, no improvement in image quality and no increase in vram usage. This means some ultra settings are missing from the game imo. You can quote me on this come final release.
I just hope the final game runs a bit better and is more optimised. As far as im concerned, mantle can't come soon enough.
Right, so walking into the midst of a flame war here but just posting this for informative purposes if other people have similar setups it would be good to hear from them.
Gigabyte WF3 7950, Clock 1050 – Memory 1300
AMD 3.10 Beta drivers
CPU AMD 8120 OC’d to 4Ghz.
Was playing Conquest Large with 64 Players for about 3 hours.
I used Radeon Pro to show on screen FPS.
Found that the FPS sat steady between 40-60(limited) all the time whilst playing. There were about three occurrences when it dropped to 25-30fps for a couple of seconds but then shot back up again (this was not consistent with smoke etc being displayed on the screen and seemed a random). I had everything set to Ultimate and 4xMSAA. Field of view was set to 90.
GPU usage 100% most of the time and maximum Vram used was 2.25Gig.
CPU usage sat steady between 60-70%
I am currently considering another 7950 for cross fire but that would mean a new power supply as well, so not overly keen on it…
Gigabyte WF3 7950, Clock 1050 – Memory 1300
AMD 3.10 Beta drivers
CPU AMD 8120 OC’d to 4Ghz.
Was playing Conquest Large with 64 Players for about 3 hours.
I used Radeon Pro to show on screen FPS.
Found that the FPS sat steady between 40-60(limited) all the time whilst playing. There were about three occurrences when it dropped to 25-30fps for a couple of seconds but then shot back up again (this was not consistent with smoke etc being displayed on the screen and seemed a random). I had everything set to Ultimate and 4xMSAA. Field of view was set to 90.
GPU usage 100% most of the time and maximum Vram used was 2.25Gig.
CPU usage sat steady between 60-70%
I am currently considering another 7950 for cross fire but that would mean a new power supply as well, so not overly keen on it…
Caporegime
That's all ive said from the start. It seems to be a common problem with the 2gb cards/690/ if you play for a decent amount of time and run High res with High settings (ultra+x4AA). 1080p/1200p obviously not included.
http://battlelog.battlefield.com/bf4/forum/threadview/2955065670086468441/
http://battlelog.battlefield.com/bf4/forum/threadview/2955065213614241730/
http://battlelog.battlefield.com/bf4/forum/threadview/2955065670086042871/
http://battlelog.battlefield.com/bf4/forum/threadview/2955064762703520265/
Soldato
Caporegime
I was just responding to questions. I'll be quiet now.
LOL
Soldato
After all this hoo-haa'ing about specs and benching, i think I'm comfortable my setup will be fine.
Caporegime
Just to show what a 3930K at 4.625Ghz is doing on a 64 man server (about 50 players). 70% on all 6 cores. Good work Dice and nice to see all the cores being pushed.
Maybe the trusty I5 is past its sell by date for this game?
Caporegime
Greg do you find HT causes any performance problems? My friend had to disable it on his 6 core sandy otherwise the game was stuttery. Once it was off it was smooth as silk.
I use HT fine on my i7 though so maybe its just a 6 core issue.
I use HT fine on my i7 though so maybe its just a 6 core issue.
Caporegime
Not at all Matt, the game feels silky in honesty.
Soldato
How much of that optimisation is due to those little boxes coming out in a month or two lol. Good to see though...
Soldato
Sorry rusty but that is not absolute.
I remember exceeding the Vram on single 3870 512MB in Race Driver Grid, on a single card @ 2560x1600 when the game first came out, it was less than 1 FPS on some settings and that was just on the menu screen.
When i added CF the FPS jumped to 25fps, when i added a 3rd GPU the FPS jumped to 70fps and the only reason why the scaling went like that is because every added card was giving the other card/cards more time to Refresh the Vram before its turn came up, but even that has limitations because when doing the 24 hour race as it got darker and darker the fps drops steadily as the Vram requirements got even higher until it stopped at 30fps.
So i added a 4th GPU my fps went to 90fps but as i got to the 24 hour race as it got darker and darker the fps drops steadily again as the Vram requirements got further and further past what i had until it stopped at 30fps again.
I just dropped 1 setting one notch in the game menu and even a single card would do 45fps avg.
If i had 3-4 way GPU from the offset i would not have been any the wiser that Vram was the cause and could have assumed it was just that the game needs more grunt than what i had when it came to the 24 hour race at night.
Now DMC3 again when i exceeding the Vram on some parts of the game it would stutter slightly, more GPUs gave me more fps but it would not stop the stuttering until i dropped the AA one notch, but no issues with cards with more Vram.
Caporegime
No difference in performance on smoothness with it on or off? Would be interesting to see how the usage is spread across the 12 threads.
Was this in the beta or bf3? As iirc early on in bf3's life one of the big stutter fixes was to kill ht.
Going by kaaps usage graphs bf4 like bf3 is topping out at 8 threads.
Caporegime
This was on Battlefield 4 pgi.
Soldato
Matt I've yet to play it on my Ivy-E but I didn't have a problem on my 3960X @ 4.8 with BF3 having stuttering as far as I can remember
Caporegime
No difference in performance on smoothness with it on or off? Would be interesting to see how the usage is spread across the 12 threads.
I will have a play at the weekend Matt but just CBA at the mo. Maybe someone else wants to do this? It is nice getting back to gaming again and not all that benching malarky
Caporegime
Thanks, maybe just a gremlin in his system then.
I do remember seeing a benchmark once (ill get him to dig it up for me) it showed faster performance in BF3 with HT off than with it ON when a 6 core cpu is used. The minimums were much higher with HT off. I always found that strange being an i7 2700k user myself where HT makes a big difference.