Review: Dying Light graphics performance

system specs directly effect how games run. a 8 core hyper threaded i7????...the game running on a ssd!!???...ludicrous amounts of the fastest ram on earth? that is not just testing the gpu. the gpu is not running 100% of the game. how the card runs in a bleeding edge pc, does not tell me how the card runs in the real world.

im actually kind of upset they are using a i7 and not a xenon. i mean...why stop at stupid specs. they could of went 1 more step up!!!

if you dont have the system used in those tests, all the numbers mean NOTHING to you.

my point was, if your gonna test 30 video cards...hell test 30 diff pc configurations as well..but AT BARE MINIMUM test the minimum and recommended specs also.

i dont have a mediocre pc. my pc exceeds the recommended requirements for every single game out to date. but i promise you, i run the same tests as this reviewer, i dont get close to him,...cpu/ram makes big differences.

second point is what good does it do to know how fast a super computer runs something when i dont have a super computer? i am simply saying...we need real world tests,more centered on the specs the manuf. gives... or more of a variety in general

i will concede that they give the GPU every possible advantage..but the results are inapplicable to the rest of the world

 

Must be a typo or mistake re the 770. Cant be on par with Titan @ 1920x1080.

 

You don't seem to grasp at the idea of GPU benchmark. This is not a system benchmark. The system is used to show the difference in horsepower of the GPU while trying to eliminate the other bottlenecks.

 

I don't know if he is just trolling now, because unless he understands what bottlenecking is, his posts are quite pointless aside from him expressing his anger at a well controlled testing environment of different GPUs

 

Also faster than the 780.

 

If you're testing the performance of any one given element, you have to make sure that no other factor can influence the result. To counter this you use any element that could interfere in an excessive state.

If I were testing to see how substance X concentration effected a given reaction, I would add all other substances in excess and modulate X, if I didn't then any of the other substances could be effecting my results.


On another note this game annoyed me with its total lack of personality and lack of decisions as to whether it was a sweet freerunning game or a dead island clone ... also weapons are stupidly unbalanced.

 

i replied better in the other thread but i dont wanna double post, but it feels more like a game benchmark to me, and thats why i mentioned min and recommended system specs. and what im saying isnt just for just this site (love the site btw long time visitor 1st time registered) but for all sites that do this. i dont dispute the points made here, but i do offer alternative viewpoints

and if i was spoiled id have a super computer. i work very hard for everything i have.

last..sarcasm is just jokes. dont be so uptight, smile some . i didnt use any profanities. i did stupid censorship stuff. but war child is right. stupid was a poor choice of words, i shouldnt have been disrespectful because i disagreed.(nor did i intend that, just loose tongue) the review was very well written, i just dont agree with the scope of it, as i am looking at the game aspect of it, rather then just measuring pure raw HP. i should have named my post more specifically to games rather then hardware in general

 

Wonder how the game plays at 1440p with 970SLI since the VRAM usage is greater than 3.5GB
And does it help to run the game on an i7 instead of i5?

 

Good review, nice indication what my system would get. Considering my cpu is clocked to 4.8. Oh and bryguy gtfo.

 

im wondering which 690 is correct on the 1920x1080 run and what the card is actually supposed to be. referencing the 2560x1440 below it, i cant figure out the card.

 

heh...from people telling me the control/bottleneck idea...i can say i get it.
the way i read the review seemed to be a specific game review more than a GPU review and its my mistake for reading it that way. i was prolly biased that way by coming from other sites reviewing the game who also shared their specs.
since you guys have told me this though i started thinking and googling...is my i5 bottlenecking my gpu? and after doing that, the sites ive seen thus far say no, even in SLI. the most reputable to me being toms hardware. (others ive never visited before)
since you guys are schooling me at what point does a bottleneck happen??
i apologize if i have offended you..its late, and if i was looking at this as purely a hardware review i wouldnt of said that, but since it was a specific game, and not 3dmark or whatever, i went a different way with it. but like i said b4, i shouldnt have been disrespectful and im sorry for that.

2nd question since the reviewer is here, did you use the dying light nexusmod launcher? or was you vanilla?

 

Good review, does the HBAO+ make that much of a difference on AMD ?, as ive always had that on, and get anywhere from 60-70/80+, Everything maxed, view distance half-way, v-sync off, due to the v-sync bug its got.

EDIT: Just tried it with the HBAO+ off, its only about 2/3fps difference.

 

from my understanding SLI does not double your vram as it ideally just alternates rendered frames between cards...so 8gbs in sli is really still just 3.5gb (for 970s +512mb or....7168/2+1024/2)=each card. and if it impacted you would depend on view distance, resolution, and texture quality. on my 970 at 1080p single card with view distance all the way down, it uses 3.4gb of vram. i should note i have not tested dying light since the new nvidia driver came out for evolve. everyone is sayin once u pass 3.5gb any 970 gets crazy slow, so to run that res, u will want to change the texture quality 1 notch down, or draw distance...etc...nvidia geforce experience is high on crack and will suggest i run brand new games at 4k resolution and scale down, so........dont suggest that either....will report back with screen caps of my results tmrw but riot can tell you.....dying light dont even know what sli is...so your biggest concern is staying under 3.5gb vram...unless newest driver changed that,(doubtful) but im sure that would be such big news it would take over my tv like a amber alert..they also say the 970 problem is hardcoated (aka laser cut) which is practically unfixable. i plan on getting another 970....but not until i cant run everything on earth at 1080p in psycho quality......but by then,..there will be 8-12gb cards on market im sure..and if that happens deff goin that route. if u wanna see what sli 970s do, check out crysis 3, or unreal tech demos, or run 3 displays in nvidia surround...very taxing. dying light is a POS though, and only worth 20$(generous only cuz multiplayer) of its 60$ price tag...so dont judge your vid cards off a broken game..

 

I have seen my 2x 970's go upwards of 3.9GB and never once had any stutter, or massive drop in frame rate in any game I play. The 970's have 4GB, not 3.5GB + 512MB they have 4GB and can access the whole 4GB. Just because they are in different pools and one suffers from lower speeds does not mean the card can not access that 512MB when it clearly does on my end.

This game was horribly coded and released with performance breaking bugs. Like a CORE 1 99% bug, multiple memory leaks, and botched SLI and Crossfire performance.

It has since been patched and performance is a locked 60fps for me completely maxed out at 1080p using DSR to 2k resolution.

 

Remember the game also runs on consoles with a mere 1.6GHz and 1.8GHz 4 core (8 module) APU

 

I see you have 16 GB of RAM.
What is your max system RAM usage while playing?

 

With 0% view distance

and my god does the 780 ever suck ass at stock