Arma 3
LoTekRabbit 2013 年 9 月 12 日 下午 1:17
ARMA 3 benchmark
Here is an official benchmark for the game.

http://www.pcgameshardware.de/Arma-3-PC-236876/Specials/Arma-3-Technik-Test-1087829/

A Titan on Extreme settings gets 19 frames. On high settings it gets 30 frames. The resolution is at 1080P. So I don't get how some of you with lower specs can claim 50+ frames on high or maxed out settings.

PS: This is just to inform people of what benchmark sites are getting. So those like me who are curious about performance can know what sites are reporting =)
最后由 LoTekRabbit 编辑于; 2013 年 9 月 12 日 下午 2:40
< >
正在显示第 31 - 45 条,共 51 条留言
Dj Otacon 2013 年 9 月 12 日 下午 1:58 
引用自 ColinLafferty
I have basically the same specs as most and i get 60fps no drop on singleplayer with vync on and 30ish on multiplayer all maxed out. View distance doesnt matter in multiplayer as all servers default to 1600m even if you have set this way higher in your own personal settings.~

I7 3770k 4.7ghz
Asus GTX titan x2
32gb ripkjaw 2100mhz
Samsung 240GB SSD (win8)
Corsair 120GB SSD (Arma3 + Rome )
2 x 1TB Black Caviar
Water Cooled

Benchmarks are never accurate and i have played from alpha right to final release. Only problems i get is some servers are unstable due to many factors which i consider and still need to turn a few settings down but get way more stability and fps than most.

But your bench is not valid because you dont have a web where a non-customer of the game can't see the real bench...

/ironic
Veristä mössöä 2013 年 9 月 12 日 下午 1:59 
There are couple of things why that benchmark is biased:
1. There is no official benchmarking way to arma.
2. They should least share the benchmarking map, so users could repeat the tests by them selves.
3. You can get every game lag and crack if you put numerous and numerous objects and units to map. We can't really know if the map they tested was ridiculous and didn't resemble real game situation at all. Check again point 2.
最后由 Veristä mössöä 编辑于; 2013 年 9 月 12 日 下午 2:00
Dj Otacon 2013 年 9 月 12 日 下午 2:00 
引用自 TenClub
引用自 steveyos

40+ in dayz and that stupid wasteland mode where nothing ever happens. What mode has lots of real and AI players that I could test? I could never figure arma multiplayer out lol.

Naah it's fine I don't want you to have to do all that. Thanks though!

I'll just wait for some more tests and patches/drivers.


Well done you have win a gift from the "premium companies". Dont waste your time and get your reward!!!...

/sarcasm
最后由 Dj Otacon 编辑于; 2013 年 9 月 12 日 下午 2:00
LoTekRabbit 2013 年 9 月 12 日 下午 2:00 
引用自 Dj Otacon
引用自 TenClub

Naah it's fine I don't want you to have to do all that. Thanks though!

I'll just wait for some more tests and patches/drivers.


Well done you have win a gift from the "premium companys". Dont waste your time and get your reward!!!...

/sarcasm

What the hell is your problem? Did I insult your family? What are you even trying to say. None of that made any sense whatsoever.

引用自 ontelo
There are couple of things why that benchmark is biased:
1. There is no official benchmarking way to arma.
2. They should least share the benchmarking map, so users could repeat the tests by them selves.
3. You can get every game lag and crack if you put numerous and numerous objects and units to map. We can't really know if the map they tested was ridiculous and didn't resemble real game situation at all. Check again point 2.

You don't need an official benchmark to test a game. You do multiple runthroughs of whatever it is you are doing and grab the average frames from there.

The point is to test the computer to the point where it is stressed out because you want the ideal framerate during those situations right?
最后由 LoTekRabbit 编辑于; 2013 年 9 月 12 日 下午 2:05
TheScotsman 2013 年 9 月 12 日 下午 2:01 
引用自 Dj Otacon
But your bench is not valid because you dont have a web where a non-customer of the game can't see the real bench...

/ironic

Actually i have a youtube channel and a few arma3 videos both machinimas and gameplay videos so i would count that as a benchmark in my own right. You can visually see the smoothness of gameplay and this is even in intense fire fights.
最后由 TheScotsman 编辑于; 2013 年 9 月 12 日 下午 2:03
LoTekRabbit 2013 年 9 月 12 日 下午 2:08 
引用自 ColinLafferty
引用自 Dj Otacon
But your bench is not valid because you dont have a web where a non-customer of the game can't see the real bench...

/ironic

Actually i have a youtube channel and a few arma3 videos both machinimas and gameplay videos so i would count that as a benchmark in my own right. You can visually see the smoothness of gameplay and this is even in intense fire fights.

To be fair you can find Crysis 3 videos on youtube with a GTX 460 on Ultra settings and it'll look smooth when in fact actually playing it is far from smooth.

Not saying this is you but I'm just giving an example. You are also using two Titans and I'd imagine this game makes good use of SLI/Crossfire.


最后由 LoTekRabbit 编辑于; 2013 年 9 月 12 日 下午 2:09
TheScotsman 2013 年 9 月 12 日 下午 2:12 
Trust me this game doesnt even know what sli means it fails in that department so im really utilising one titan. Yeh i suppose your correct but i believe this game should be labeled "need a monster pc to enjoy it" because on single player everyone can play it without many problems but multiplayer really takes the cake.

Also running from an ssd actually helps an awful lot in my experience as wel as ram disks (i have only seen videos) especially will load times etc. Obviously it wont help with fps if textures are been pushed around with the gpu but if they are being pulled directly from the hard drive to the gpu then thats when an ssd can really help.
最后由 TheScotsman 编辑于; 2013 年 9 月 12 日 下午 2:18
Veristä mössöä 2013 年 9 月 12 日 下午 2:17 
引用自 TenClub

引用自 ontelo
There are couple of things why that benchmark is biased:
1. There is no official benchmarking way to arma.
2. They should least share the benchmarking map, so users could repeat the tests by them selves.
3. You can get every game lag and crack if you put numerous and numerous objects and units to map. We can't really know if the map they tested was ridiculous and didn't resemble real game situation at all. Check again point 2.

You don't need an official benchmark to test a game. You do multiple runthroughs of whatever it is you are doing and grab the average frames from there.

The point is to test the computer to the point where it is stressed out because you want the ideal framerate during those situations right?

So everybody can say anything if they just put numbers on website and don't tell in which situation they got them?

Its down to basic research principles. If people can't mimic/repeat your tests, they are invalid.

For. ex I could say that BF3 runs like ♥♥♥♥. But I won't tell to the public that I made my own map, which is big as Altis and has thousands and thousands of units, action & explosions going on.

That is why it would great to have official benchmark so all people could test with same requirements. Or least that site could share the testmap.
最后由 Veristä mössöä 编辑于; 2013 年 9 月 12 日 下午 2:20
LoTekRabbit 2013 年 9 月 12 日 下午 2:21 
引用自 ontelo

So everybody can say anything if they just put numbers on website and don't tell in which situation they got them?

Its down to basic research principles. If people can't mimic/repeat your tests, they are invalid.

For. ex I could say that BF3 runs like ♥♥♥♥. But I won't tell to the public that I made my own map, which is big as Altis and has thousands of units, actions & explosions going on.

That is why it would great to have official benchmark so all people could test with same requirements. Or least that site could share the testmap.

No, you let people know what level you used, run whatever section you're running a couple of times and post your minimum, average, and maximum framerate.

And by the way, I don't know if you translated the site and actually read it but:

"For our benchmark scene, we choose this time an airfield on the main island Altis far more challenging. There are some NPCs running around, planes and helicopters flying past us..."

It goes on and on so you did not actually look through the article with a translator. That would have really helped before you accused them of all that.

They've been doing this for years they know what they're doing and they have no reason to be dishonest.

引用自 ColinLafferty
Trust me this game doesnt even know what sli means it fails in that department so im really utilising one titan. Yeh i suppose your correct but i believe this game should be labeled "need a monster pc to enjoy it" because on single player everyone can play it without many problems but multiplayer really takes the cake.

Also running from an ssd actually helps an awful lot in my experience as wel as ram disks (i have only seen videos) especially will load times etc. Obviously it wont help with fps if textures are been pushed around with the gpu but if they are being pulled directly from the hard drive to the gpu then thats when an ssd can really help.

Damn, is this becoming a trend? High end games not utilizing SLI/Crossfire? Company of Heroes 2, Rome 2, now ARMA 3. I hope they get some patches and profiles out so you can use both Titans bud.
最后由 LoTekRabbit 编辑于; 2013 年 9 月 12 日 下午 2:32
Deaths 2013 年 9 月 12 日 下午 4:37 
I get 80 + frames playing single player campaign flying the jets while being shot at by the submarine ....;0
oxmox 2013 年 9 月 12 日 下午 4:38 
引用自 BIXBY SNYDER
Well, to be fair, I've seen this game have wildly different performance on almost identical systems.

Are you guys really testing the FPS with Arma 3 or do you play with certain player made modifications i.e. Wasteland, Life-servers (rpg) or CTI.
最后由 oxmox 编辑于; 2013 年 9 月 12 日 下午 4:39
Алзксі Венс 2013 年 9 月 12 日 下午 5:25 


引用自 TenClub
Damn, is this becoming a trend? High end games not utilizing SLI/Crossfire? Company of Heroes 2, Rome 2, now ARMA 3. I hope they get some patches and profiles out so you can use both Titans bud.
SLI/Crossfire computers are such a small portion of the market and it's so expensive to cater to them that most devs couldn't really give a damn if it utilizes both of your GPUs.
最后由 Алзксі Венс 编辑于; 2013 年 9 月 12 日 下午 5:25
Event Horizon 2013 年 9 月 12 日 下午 5:35 
After playing wasteland,sandbox,co op type servers I've gotten an avg framrate of 23 FPS this is on Ultra settings. I didnt spend much time on each server around 15-20 minutes.

From what little I played so far the general FPS I am getting is around 20ish never anything over 30, maybe 32-33 max with fraps off.

i7 4770k, single GTX 780, samsung 840 pro.
最后由 Event Horizon 编辑于; 2013 年 9 月 12 日 下午 5:36
ISSA_Humps 2013 年 9 月 12 日 下午 5:49 
i have a i7-3770k oc'd 4 ghz 16 gb ram 2133 mhz and a samsung ssd these setting in the set launch options always help i used them in both arma 2 and arma 3

-nosplash -cpucount=4 -exthreads=7 -nofilepatching -nopause -maxmem=2047 -maxvram=2047


i get 80 fps now with maxed out settings and view distance at 3800

hope this helps
LoTekRabbit 2013 年 9 月 12 日 下午 8:05 
引用自 That Guy
引用自 TenClub
Damn, is this becoming a trend? High end games not utilizing SLI/Crossfire? Company of Heroes 2, Rome 2, now ARMA 3. I hope they get some patches and profiles out so you can use both Titans bud.
SLI/Crossfire computers are such a small portion of the market and it's so expensive to cater to them that most devs couldn't really give a damn if it utilizes both of your GPUs.

It's 2013 and the vast majority of devs cater to SLI and Crossfire. You're telling me that multi platform devs that release console ports are able to cater to SLI/Crossfire users but developers that make PC exclusives can't?
< >
正在显示第 31 - 45 条,共 51 条留言
每页显示数: 1530 50

发帖日期: 2013 年 9 月 12 日 下午 1:17
回复数: 51