Caporegime
Using vsync or an fps limiter? If so disable it.
-
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
Using vsync or an fps limiter? If so disable it.
Associate
I don't use Vsync and I'm not limiting my fps
is this Nvidia's power limit control, in the control panel I use "Max performance" Thanks Matt
Caporegime
I had the same problems with a 2500K at 4.6 and SLI 680's in BF3. I could never get the GPU's at 99% usage. I switched out for a 3930K and at 4.6 on that, both GPU's were at 99% GPU usage.
Caporegime
I don't use Vsync and I'm not limiting my fps
Thanks Matt
No idea what that is im afraid mate.
Soldato
No image quality difference really at such low res between x2 and x4. The only difference is the marginally better gpu usage. (while still not perfect - should be 99% solid) 1440P with FXAA will look 99 times better than 1080p with x4 AA.
7990 still overkill for 1080p imo. If you have a 120hz screen though ill grant its not overkill. So consider this a compromise of sorts.
While I kind of agree that there isn't a huge difference between x2 and x4, I will have to dispute your fxaa @1440 claim with great gusto
Caporegime
While I kind of agree that there isn't a huge difference between x2 and x4, I will have to dispute your fxaa @1440 claim with great gusto
I have a 1440p and a 1080p screen right next to each other and i know which one looks better, in my opinion at least.
Soldato
Are we just talking about TR here?
Caporegime
I can't tell the difference either Pete. Even with different resolutions and amounts of AA, the 1440P looks the same as the 1080P screen to me. Deeper colours on the 1440 but that is it.
Are you running 2 780s there Setter or have you still got your 670s?
Caporegime
I've yet to see a game on my pc look better at 1080P with AA than 1440P with some form of FXAA.
Caporegime
Still on the 670's. Should have my 780 tomorrow.
Soldato
That's fair enough. From what I've seen, everyone seems to have a different opinion.
Some see a massive difference, some see none.
At 1080, I see a noticeable difference between x2 and x4 in BF4/Crysis 3/Metro. Batman games at x8, and a slight difference between high/extreme in sleeping dogs for example.
Tomb raider doesn't seem to show a really noticeable difference, but I'd put that down to most of it being soft edged jungle stuff rather than hard edged city type stuff.
With games like C3, I still maintain that a 7990/79xxCF isn't overkill at 1080. At lower AA/settings (even high/x2AA) there is very noticeable aliasing in interior locations compared to the solidity at very high/x4. A 7990 shines here.
Hey there, don't mind me, but I thought I might drop some actual BF4 benchmarks up in this bizzatch
1080p.
Shanghai 64 All Ultra/4xMSAA with FXAA disabled for all three tests and all cards set to 1000/1500 with a 2600K @4.7Ghz.
7990
Frames, Time (ms), Min, Max, Avg
69256, 707828, 57, 174, 97.843
Power usage at the wall: 480W. Not sure if it peaked any higher.
Temps/GPU usage:
7990+7970 Trifire
Frames, Time (ms), Min, Max, Avg
57386, 582875, 57, 196, 98.453
Power usage at the wall: 532W average with occasional peaks up of 640W
Temps/GPU usage:
7990+7970 Trifire @200% Resolution Scale
Frames, Time (ms), Min, Max, Avg
38645, 653969, 0, 93, 59.093
Power usage: 650-680W
Temps/GPU usage:
Vehicles exploding and spectacular deaths were responsible for those low digit/zero minimums @200%.
It may look slightly more crisp than 100%, but I don't think it's worth the performance penalty.
So all in all... You're gonna need a high res screen to really take advantage of TriFire and probably a rather juicy cpu OC too.
I know, shocking news, right?
Thats a real eye opener Pete. So whats the issue with the mins?
The mins show that there is a coding bug becuase no way should a single 7990 have the same mins as a 7990+7970, something isnt right. (BF4 wise)
Caporegime
Tried it with 200% res @ 1080P all Ultra / Post AA low / 0 MSAA.
7870XT, yeah, it didn't go to well didn't think it would.
7870XT, yeah, it didn't go to well
didn't think it would.
Caporegime
Its called a cpu bottleneck and would be interesting to see Petey run this test again via the mantle api. Sadly that won't happen though. If my 2700k @5ghz bottlenecks my stock 7990 at 1080p, you can be sure Pete is getting a bottleneck with a slower clocked cpu and another 7970 in there to boot.
Soldato
Not 100% sure, but three cards are blatantly held back by the resolution and the CPU to a degree. They'll hover at around 50-60% usage though at least they all match each other.
While I don't personally think a 7990 is wasted in demanding games at this res, trifire definitely is.
Those results arent that far off Petes 7990, 7990+7970 bench marks above.
Maybe changing Gpu wasn't necessary Setter. (Maybe the game is so badly coded for sli/crossfire/trifire performance is signigicantly pegged.
I'm also sure that i had mins of 52 with a single 7950 (1200mhz) in Seige!!!
Last night the game was lagging, mins fps's 49 (in crossfire) and loads of crashes and BS-OD.
Caporegime
Setter is only using x2AA and his cpu is clocker faster than Pete's.
That makes sense.
But last night probably wasn't a good time to benchmark because virtually all the servers I joined were running slow and unstable.
I also saw very low fps mins so maybe there were issue's server side because the game was freezing and crashing virtually every map I joined.
Caporegime
Yes server issues will cause low fps. Someone ddos'ing a server will cause gpu usage to drop as well resulting in low fps.