Dying Light: Performance Analysis And Benchmarks

[JonnyDough]

"Never before have I stopped playing a game to question my morals and what I felt right about doing."

Surprising, there are many games today that push this envelope and it's not one we should really be pushing...not for money and not for entertainment. How polluted have we become? If it makes you feel sick that's a good thing.

We should all question our morals with games like this. The only reason the developers hold back these days is because killing children is still considered taboo. Even the notion of it is a terrible sign that we have degraded as a society. We no longer seem to hold ourselves and each other in high regard and at one point is what common to want a better life for our children and to put them first. Ask any cop, juvenile court officer, or social worker and they will cry out societies woes to you. This is disturbing and it should be, although probably not enough for many today.

 

[sephirotic]

This is a perfect game to show how the internet neurotic machine is full of bs regarding the slow 512 VRAM on the GTX 970. I have one and the game does not stutters at all with over 3.5gb of ram being used. The time variance compared to the GTX 980 and with the AMD cards (which are also) also proves that.
Not that I was happy about the false advertising, but is good to have this article around everytime a missinformed fool says he doesn´t want to buy the GTX 970 "because it only has 3.5gb of ram".

 

[silverblue]

AMD CPUs are also scoring the same regardless of clock speed. Something's up.

 

[Eggz]

Strange, how is my crossfire 4GB 290X set up getting steady 60 FPS with very high textures @ 4K while the author is getting upper 20's?

Either the author isn't smart enough to know how to turn on AFR mode on his crossfire profiles (not a good trait for PC reviewer), or he intentionally tried to get horrible results on purpose (also not a good trait for a PC reviewer).

Any chance of an explanation for this?

Calling crock!

I think you're the only person on the internet getting a min of 60 fps in a modern and demanding title while running 4K on a dual card setup. Your setting also sound different than his.

Plus, the reviewer didn't even use CF or SLI - only single cards. I don't even think there's CF and SLI profiles out for this game yet.

 

[Cryio]

Techspot showed that the game performs indentical to Intel levels with an FX 9590 at 5 GHz. Tom's bench must have gone haywire somewhere.

 

[killerchickens]

how well would my i7 960 4.2ghz do? gtx 970

 

[killerchickens]

I want to see an Intel Pentium G3258 bench marked overclocked of course.lol

 

[Eggz]

Benchmarks of this game (and any without a built in bench) all suffer from the same consistency problem; namely, they sample only a small part of the game. What about when you're looking straight up at the sky? Can you really say you get 200 fps? Of course not. That's cherry picking an easy scene. The same goes for benching the final boss battle. Can you say in that case that your minimum dropped down into the 30s with you GTX 980 / 290X? In that case, yes, you can! That's actually what minimum's are for - a worst case scenario. And that's actually what Tom's did here that TechSpot didn't do over there.

TechSpot ran a bench inside a building, not outdoors, where it's extremely easy for the hardware to render frames because the draw distance is so short (i.e. less than half the work for the hardware to do). They even say to read the benchmarks as if they had the draw distance turned all the way down to 50%. Compare the following statements in each article:

[M]any reviews show performance to be around half what we saw. The issue seemed to be the draw distance, which doesn't have an impact inside, even when looking out over the landscape as we did. . . .

It isn’t until you hit ground level and walk outside that the frame rates take a huge hit with the draw distance set to 100%. Backing the draw distance off to 50% restores performance and gamers will see the numbers we showed in this article.

. . . For now although we tested with 100% draw distance, the results should be treated as though the distance was set to 50%.

Tom's didn't do that. They actually picked a challenging part of the game to render, which gives a better assessment of what to expect in terms of potential system draw.

Our custom benchmark took a while to concoct. We settled on a spot that requires a fair bit of running from the home base, since there is no fast travel option. It begins in front of an abandoned school and continues to a bridge with a car that is on fire. This creates a dip in performance and shows the game's beautiful fire effects.

So this all shows two things to me: (1) AMD, Intel, and Nvidia all perform similarly when things aren't that hard to render, but (2) Nvidia and Intel do better when things get challenging. Maybe a driver update or game patch will change that, but for now that's what we're seeing.

 

[Sheik_Derez]

The benchmarks clearly shows a very poor multicore utilization. Just wait for Directx12 and we will see 9590 surpass i3 and i5 with ease.

 

[killerchickens]

How does this game run on the consoles? 8 slow amd threads this performance makes no sense.

 

[iam2thecrowe]

wow what a horribly optimized title. If you have an amd cpu and gpu you are f*cked. The most confusing thing is how the fx-4170 outperforms the 9590. I cant even begin to comprehend how that is possible, unless the game devs purposely gimped the game engine if it detects more than 4 cores.
Also the gtx970 got slightly higher fps at 1440p than 1080p? WTF? not going to play this game out of sheer principle that the optimization is poo.

I'd actually like to see gpu and cpu usage figures of the poor performing amd bunch. Are we able to get those??

 

[iam2thecrowe]

Typical Nvidia Gameworks title. Anyone remember Metro: Last Light? Unplayable on AMD cards until 4A issued a game update a few months later. Can't make a card that competes? Pay off the game developers.

the silly thing is, their cards DO compete.

 

[Dustin Mock]

@Eggman:

295x2 is a crossfire 290X set-up which was one of the cards tested.

Would you like me to Fraps, or will screenshots suffice?

 

[Eggz]

@Eggman:

295x2 is a crossfire 290X set-up which was one of the cards tested.

Would you like me to Fraps, or will screenshots suffice?

I'm familiar with 295x2. Would your fraps/sc show 60+ fps @ 4k or xf? If fps, show outdoor scene.

 

[DecafTable]

Game looks interesting but will skip because of poor optimization for AMD cards. I have no issues playing Alien Isolation maxed no drops whatsoever. Would be nice if more developers could get with the program.

 

[Dustin Mock]

http://img.pixady.com/2015/03/386495_4k.jpg

http://img.pixady.com/2015/03/176905_4k_2.jpg

http://img.pixady.com/2015/03/515179_4k_3.jpg

http://img.pixady.com/2015/03/158076_4k_4.jpg

http://img.pixady.com/2015/03/292915_4k_5.jpg

http://img.pixady.com/2015/03/873384_4k_6.jpg

http://img.pixady.com/2015/03/580997_screenshot_9.png

Here you go on screenies I just took a sec ago, don't have fraps installed at the moment but I'll get it in a sec

 

[Dustin Mock]

@eggman

Been a while since I used fraps. Here is my log though from running around outside right quick:

2015-03-06 18:57:33 - DyingLightGame
Frames: 1203 - Time: 22094ms - Avg: 54.449 - Min: 43 - Max: 63

 

[Eggz]

I see. Did you fraps in the same area your took sc? The math on that fraps readout says you sampled about 27 sec, a little short to conclude stuff. Already know there's good performance in easy areas (see longer post above). Maybe you can find the same run by the fire Tom's described in the article (quoted above) for better comparison.

 

[Dustin Mock]

No those aren't from the same I just loaded the game, ran outside tried to grab screenshots of action scenes quick real quick. The log I did real quick running around the same area outside after that. Just made a quick video running outside to show that it's actually pretty smooth. unfortunately I know nothing about video editing so I just took the fraps video and hit upload on youtube.

http://youtu.be/i4ncAJoq8cM

 

[Dustin Mock]

I'll try and re-create the scene they described and see what I can pull off, although I am almost positive they didn't use AFR and just left it on the default.

http://forums.amd.com/game/messageview.cfm?catid=475&threadid=183935&STARTPAGE=2&FTVAR_FORUMVIEWTMP=Linear

When I put it to default instead of AFR my FPS drops down considerably.

 

[Eggz]

I'll try and re-create the scene they described and see what I can pull off, although I am almost positive they didn't use AFR and just left it on the default.

http://forums.amd.com/game/messageview.cfm?catid=475&threadid=183935&STARTPAGE=2&FTVAR_FORUMVIEWTMP=Linear

When I put it to default instead of AFR my FPS drops down considerably.

Hmm, if that turns out to be the case, it may be worth dropping a note to them. Wait and see, though, until the scene they used does the same stuff.

 

[iam2thecrowe]

http://www.techspot.com/review/956-dying-light-benchmarks/page5.html
Look over here, seems they have completely different results to toms. Amd gpu's are slightly behind nvidias but not like what toms recorded. And the CPU bench is completely different with the fx amd cpu's competitive with the ivy bridge i5. What is going in with your test bench toms????
They did this 'For testing we maxed out every setting with the exception of Nvidia HBAO+" which is fair enough.
and this "Post-test notes:

After further investigation we have learned a few things about Dying Lights performance with the latest v1.2 patch. The performance we found indoors is accurate, however many reviews show performance to be around half what we saw. The issue seemed to be the draw distance, which doesn’t have an impact inside, even when looking out over the landscape as we did.

Our test does include a section where the player walks to the edge of the tower and looks out over the landscape, here the frame rates actually increase significantly which was unexpected.

As it turns out the draw distance isn’t an issue here either as finer details are not displayed. It isn’t until you hit ground level and walk outside that the frame rates take a huge hit with the draw distance set to 100%. Backing the draw distance off to 50% restores performance and gamers will see the numbers we showed in this article.

It is believed that the draw distance performance at 100% can be cured via a patch and if this is the case then the performance we showed inside will be much the same outside as well. For now although we tested with 100% draw distance the results should be treated as though the distance was set to 50%."

 

[MagusALL]

Well AMD have to develop both CPU's and GPU's, so you can't expect them to operate at a pace to beat two well established hardware companies, Intel and nVidia. I think it's just simpler than people make it. If you are filthy rich would you buy a practical vehicle expecting your richness to make the Kia a Lambo? No so why are you comparing cards and introducing all these limiting agents like CPU. Just throw the best one on there, along with RAM, PSU and then benchmark. Don't expect a company that is trying to make money by developing their own units of computer hardware which is in high demand to MAX out every game. If you want to be a PC gamer man up and get yourself a i7 Intel with 4 cores and OC it, then buy a nVidia card that matches your display. There, the great mystery has been solved, your welcome.

 

[MagusALL]

you're* welcome...dammit!

 

[chimera201]

Maybe in Tom's setup , the FX CPUs got overheated and performance was throttled. That could explain why the FX 9590 performed lesser than FX 4170. Just speculating since the test setup for the CPU benchmark wasn't given.