Support NeoGAF

[Effect]

Okay I think I founds settings that work fairly well. Usually getting around the low 40s FPS wise with dips and the occasional stutter.

FX 8320 3.5Ghz
GTX 960 2GB
8GB Ram
1920 x 1080

The volumetric lighting is giving me serious issues. Turning it off keeps things more stable. I've been toying with texture filtering being 2x and 4x Anisotropic.

So far this has worked. I wish it wouldn't dip though.

However I think the game still looks pretty good.

I have to say though if I had a PS4 I really do think I might have gotten it on that. Might not look as good but I imagine things would be more stable and consistent. At least I would hope so. I just hope future patches makes things run better as I am enjoying the game.

Going to mess with lowering my resolution to see what kind of impact that has.

 

[Fledz]

Initial load time is killing me now, after 14 hours play in the North part of Prague. Timed it at 11min last time :|

Must be the 6GB RAM surely? I've bumped textures down from high to see if it speeds it up next time. If it looks shit though I'll just bite the bullet and wait for it to load, while I make a sandwich or something. I'm dreading that train ride back...

 

[icecold1983]

Heck yes it is, don't get me started with its implementation in Far Cry 4

But in direct comparison to this game's CHS, it's head and shoulder above it when implemented correctly. It just seems like a missed opportunity on so many levels, and it's apparent that Nixxes hands were tied with AMD involved. No, I don't want any one company to have a monopoly on the market (that's not healthy), but it's also obvious that AMD doesn't have nearly as good proprietary effects as Nvidia does, which Nixxes has used excellently in the past.

thats a big if considering its complete shit quite often. ac syndicate, far cry 4, gta v among others

 

[loganclaws]

Initial load time is killing me now, after 14 hours play in the North part of Prague. Timed it at 11min last time :|

Must be the 6GB RAM surely? I've bumped textures down from high to see if it speeds it up next time. If it looks shit though I'll just bite the bullet and wait for it to load, while I make a sandwich or something. I'm dreading that train ride back...

Lol 11 minutes?!?!?!?!?!?!!? You serious?

 

[MrOogieBoogie]

I can't imagine a serious patch to fix performance and mouse sensitivity doesn't release soon. Can't recall the last time a game has had such widespread outcry. Surely the devs are listening.

Okay I think I founds settings that work fairly well. Usually getting around the low 40s FPS wise with dips and the occasional stutter.

FX 8320 3.5Ghz
GTX 960 2GB
8GB Ram
1920 x 1080

The volumetric lighting is giving me serious issues. Turning it off keeps things more stable. I've been toying with texture filtering being 2x and 4x Anisotropic.

So far this has worked. I wish it wouldn't dip though.

However I think the game still looks pretty good.



I have to say though if I had a PS4 I really do think I might have gotten it on that. Might not look as good but I imagine things would be more stable and consistent. At least I would hope so. I just hope future patches makes things run better as I am enjoying the game.

Going to mess with lowering my resolution to see what kind of impact that has.

I have a very similar system (i3-2100 + GTX 960 2GB + 8GB RAM) and I think you can turn off several settings to really boost your performance with minimal visual degradation. Namely:

Level of Detail: Medium
Cloth Physics: Off
Subsurface Scattering: Off
Tessellation: Off

And perhaps setting Texture Filtering to 4x.

Surely this will make your game run better.

Do you have MSAA off as well? How about Exclusive Fullscreen? VSync?

 

[dr_rus]

I said new game, so DOOM (Vulkan), Warhammer: TW and Forza Motorsport Apex. 480 is clearly faster than 1060 in these games. It's a trend plain to see. This is more in reference to review sites that use loads of old games like BF3, Crysis 3 or GTA5 where Nvidia cards have a clear advantage though tbh.

Doom was faster on NV for two months after release and it is a known fact that it's Vulkan renderer is solely GCN optimized. Not a valid comparison point until id will optimize the renderer for NV h/w as well.

WarhammerTW isn't faster on 480 than on 1060 even in AMD's half-baked DX12 effort and 1060 is even faster than that in DX11 (no quality difference). False claim.

FM6A isn't really faster either as it depends on the benchmark - I've seen both cards being faster in different benchmarks. So it's a tie basically.

The only "trend" here to speak of is the trend of AMD GE titles being completely unoptimized for NV h/w. For each GE title there are a dozen of non-GE titles where 1060 usually wins.

 

[Fledz]

Lol 11 minutes?!?!?!?!?!?!!? You serious?

Yep. Was much better to start with, but I've been playing all day, it's hot, could be affecting the HDD. It's an old one either way. I'll deal with it lol, but it is a bit ridiculous.

 

[·feist·]

Kind of disappointed in the performance. My i5 3570k @ 4.2ghz really struggles. The RX480 is great, but the CPU really seems to bottleneck this game a ton. Especially in certain areas of the hub.

I wish more games prioritized GPU like Battlefield.

DX12 in early September might be a cure for what ails you, specially on AMD hardware.

Yeah, curious to see what it can do!

This is only one source, so hardly conclusive, though it wouldn't be the first time a benchmark showed lower CPU usage on the RX480 against comparable Maxwell/Pascal:

Joker Productions —— Deus Ex RETEST | AMD RX 480 vs Nvidia GTX 1060

If accurate this is still not the more common outcome, owning to the differences in both vendors' architectural focus and the nature of how their drivers are geared.

Without being built "from the ground up" to take full advantage of DX12/Vulkan, there is only so much improvement that can be extracted coming from DX11. Still, if Nixxes' DX12 rollout goes well we should she benefits across a range of Nvidia and AMD cards.

Benchmarks:
Bit-Tech (launch build[?]; DX11)
GameGPU (pre-release; DX11 and DX12)
PC Games Hardware (launch build and pre-release; DX11)
Tech PowerUp (launch build; DX11)

Guru3D —— Deus Ex: Mankind Divided PC GPU (DX11) performance benchmark review
https://www.guru3d.com/articles-pag...-graphics-performance-benchmark-review,1.html

Image quality settings and benchmark system
https://www.guru3d.com/articles_pag..._graphics_performance_benchmark_review,5.html

Frametime and latency performance
https://www.guru3d.com/articles-pag...-graphics-performance-benchmark-review,8.html

Above an FCAT plot of latency relative to FPS in percentiles. We use the Geforce GTX 1070 and the Radeon R9 Fury. I often get asked the question why we do not include the faster Fury X here, well FCAT is dependant of a DVI monitor output, and AMD is not implementing them any more on their reference products. Only board partner cards release DVI enabled products. Hence the R9 Fury we use is the STRIX from ASUS, as it has a proper DVI output connector. So at 50% you could consider to be the average frame-rate. The cards are nice and close and cuddly to achother. The plot is based on the first 31 seconds measured in the benchmark.

Frame Pacing / Frametime

Above a the frame-time results plot of the test run @ 2560x1440 (WQHD) performed with a GeForce GTX 1070 in 2560x1440 (WQHD). That's picture perfect rendering really.

And above the R9 Fury. The spike at the beginning sometimes can happen when you startup the FCAT sequence, this is not an error from the graphics card. Here again we see really okay frame-time behaviour overall. Remember you are looking at close to roughly 2000 rendered frames; all nicely plotted to see if anything massively weird happens. That is just not the case. And now let us overlay and combine the two:

Both graphics card manufacturers seem to have the game well under control ralative to frame pacing, this is close to excellent. Sweet results here for both brands really.

Decent frame-time performance on both vendors (assuming the different modern GPU families from both received similar attention), so that's at least one less technical hurdle that needs to be immediately addressed.

 

[CertifiedFP]

Sounds like a system issue. Doubt a patch will fix that.

Literally every other major AAA game of this year runs completely perfect on this PC, most of them on max.

This is a port issue, plain and simple.

Also I am certain this game has a memory leak going on.

 

[jorimt]

thats a big if considering its complete shit quite often. ac syndicate, far cry 4, gta v among others

Well, since the CHS in this game is basically unusable, let's say it sets a low bar. That, and PCSS isn't the only shadow softening tech available to Nvidia. What I'm saying is, stack any AMD port's enhanced/proprietary effects against Nvidia's, and there isn't even a comparison. I'm simply disappointed is all.

 

[Effect]

I can't imagine a serious patch to fix performance and mouse sensitivity doesn't release soon. Can't recall the last time a game has had such widespread outcry. Surely the devs are listening.



I have a very similar system (i3-2100 + GTX 960 2GB + 8GB RAM) and I think you can turn off several settings to really boost your performance with minimal visual degradation. Namely:

Level of Detail: Medium
Cloth Physics: Off
Subsurface Scattering: Off
Tessellation: Off

And perhaps setting Texture Filtering to 4x.

Surely this will make your game run better.

Do you have MSAA off as well? How about Exclusive Fullscreen? VSync?

Yeah I have MSSA off and vsync off too. I've gone back and forth on exclusive fullscreen. As for the other settings I didn't notice to much differnce in fps or performance. I figure if there isn't that much change or doesn't appear to be it might as well look nicer.

 

[icecold1983]

Well, since the CHS in this game is basically unusable, let's say it sets a low bar. That, and PCSS isn't the only shadow softening tech available to Nvidia. What I'm saying is, stack any AMD port's enhanced/proprietary effects against Nvidia's, and there isn't even a comparison. I'm simply disappointed is all.

hbao+ aside, i think on average AMD added effects are much more consistent and reliable than nvidias. gameworks has been absolute garbage for a couple years now.

id consider pcss unusable in the above mentioned games too. performance and visuals are both worse than the default ultra options in all cases. sometimes drastically so

just to be clear, im not saying amds effects arent crap in this title

 

[dr_rus]

hbao+ yes, pcss is just as spotty

PCSS is head and shoulders above CHS and it's definitely not "just as spotty".

https://www.youtube.com/watch?v=Jh95VU3FY40

hbao+ aside, i think on average AMD added effects are much more consistent and reliable than nvidias. gameworks has been absolute garbage for a couple years now

And this is just completely false. AMD added effects tend to look worse and run worse on AMD's own h/w as well as NV's. The time when AMD did add good code to games have ended somewhere back in early DX11 days or probably at the launch of 7970.

 

[Fledz]

So according to this, turning off V-SYNC improves load times. Sure, will give it a try tomorrow.

https://www.reddit.com/r/Deusex/comments/4ysrsz/level_loading_times_in_mankind_divided/

 

[jorimt]

hbao+ aside, i think on average AMD added effects are much more consistent and reliable than nvidias. gameworks has been absolute garbage for a couple years now.

id consider pcss unusable in the above mentioned games too. performance and visuals are both worse than the default ultra options in all cases. sometimes drastically so

I'm not going to even begin arguing with you on the point that AMD's effects are superior, simply because they don't have anything that is repeated enough to be memorable, let alone list.

However, I will say, while I don't think PCSS has always been implemented in the best way possible, we're not talking about performance impact here, were talking about whether it works or not. That, and it does work well in some games, as long as you are able to run it:
http://images.nvidia.com/geforce-com/international/comparisons/assassins-creed-syndicate/assassins-creed-syndicate-shadow-quality-interactive-comparison-004-pcss-ultra-vs-high.html
http://images.nvidia.com/geforce-com/international/comparisons/assassins-creed-syndicate/assassins-creed-syndicate-shadow-quality-interactive-comparison-002-pcss-ultra-vs-high.html

Note the increased fidelity and draw distance.

Now compare it to this:
http://www.neogaf.com/forum/showpost.php?p=214763970&postcount=783

If you're going to lose 12-15 fps, it better be worth it.

EDIT: Corrected first link. It was a duplicate of the second one.

 

[Courage]

Umm... heres me grasping for words.

I fiddled with the graphics options for couple of hours and I did manage to make the game look and roll fairly nicely. I have my computer details and benchmark results below and more info about the actual graphics optins in the information slab on Youtube:

THE COMPUTER

Resolution: 2560 x 1440
i74970k (not OCd)
MSI 1070 Gaming X
12GB
SSD

IN-GAME BENCHMARK TOOL RESULTS:

Average: 45.9 FPS
Minimum: 38.1 FPS
Maximum: 56.9 FPS

I aimed to have the minimum as close as 40 FPS that I could, but I didn't want to drop all the most delicious options. There were few surprises, like how the game ran better with Texture Resolution on ULTRA rather than on HIGH or MEDIUM. Also, the brutality of the MSAA caught me by a complete surprise, but fortunately the TAA is very nice.

Overall, I'm pretty happy with how I got it looking. Things look like they belong there and nothing sticks out like a sore thumb... as long as you understand to keep away from the Sharpening filter

Wow and that's with Contact Hardening Shadows on Ultra? Another reason to upgrade from my 970 to a 1070 and invest in a 1440p monitor.

 

[Unknown Soldier]

I think Sombra is invading my game



Hope it's a software issue not a hardware one. First time something bad has happened so far in a game with my 1070

I'm getting this bug too on a 1080. You can force it to happen by manually saving your game in certain areas.

 

[icecold1983]

PCSS is head and shoulders above CHS and it's definitely not "just as spotty".

https://www.youtube.com/watch?v=Jh95VU3FY40



And this is just completely false. AMD added effects tend to look worse and run worse on AMD's own h/w as well as NV's. The time when AMD did add good code to games have ended somewhere back in early DX11 days or probably at the launch of 7970.

video doesnt show the full picture. with pcss every shadow cascade beyond the first 1 or 2 look like old ps3 pcf shadows and many distant shadows display striping artifacts if they even render at all. none of this happens on the rockstar coded softest option.

 

[dr_rus]

video doesnt show the full picture. with pcss every shadow cascade beyond the first 1 or 2 look like old ps3 pcf shadows and many distant shadows display striping artifacts if they even render at all. none of this happens on the rockstar coded softest option.

Oh yeah, so much better: https://www.youtube.com/watch?v=QCCdMaLerMg
Or is this a bad video as well?

 

[Intheflorsh]

I'm getting a 10-20 second freeze everytime I open the menu or pause the game, any workaround?

 

[texhnolyze]

Have there been benchmarks and/or reports on vram usage for each texture settings?

Okay I think I founds settings that work fairly well. Usually getting around the low 40s FPS wise with dips and the occasional stutter.

FX 8320 3.5Ghz
GTX 960 2GB
8GB Ram
1920 x 1080

The volumetric lighting is giving me serious issues. Turning it off keeps things more stable. I've been toying with texture filtering being 2x and 4x Anisotropic.

So far this has worked. I wish it wouldn't dip though.

However I think the game still looks pretty good.

I have to say though if I had a PS4 I really do think I might have gotten it on that. Might not look as good but I imagine things would be more stable and consistent. At least I would hope so. I just hope future patches makes things run better as I am enjoying the game.

Going to mess with lowering my resolution to see what kind of impact that has.

You have a 2gb video card, you should use medium texture instead of high. The dips might be caused by the lack of vram. I also have a 2gb card and been using medium texture, capped at 30fps and everything's fine.

I'm getting a 10-20 second freeze everytime I open the menu or pause the game, any workaround?

I'm not sure, but it seems like ram issue. I also have similar issue if I open chrome on my tabs while playing.

 

[Collateral22]

Getting a lot of random black screens and crashes. Also really horrible load times, like over 1 minute. That the downside to only 8GB RAM?

 

[Rabbit Clock]

This runs poorly on just about any settings I try. Long hangups every time I open up menus. Random FPS drops.

I have a 970 so it should run decently at least but no matter what I lower it always ends up around 52 FPS. On higher settings (no MSAA) it hovers between 50-55 fps and on the lowest settings it hovers around 50-55. I don't even care that much what it looks like.. I just want a solid frame rate. I guess I'll wait for a patch or something.

 

[texhnolyze]

Getting a lot of random black screens and crashes. Also really horrible load times, like over 1 minute. That the downside to only 8GB RAM?

After 4 hours in, no black screens nor crashes so far.

Loading time is indeed long for me as well, but I guess it has something to do with my 6 year old HDD & CPU.

 

[icecold1983]

I'm not going to even begin arguing with you on the point that AMD's effects are superior, simply because they don't have anything that is repeated enough to be memorable, let alone list.

However, I will say, while I don't think PCSS has always been implemented in the best way possible, we're not talking about performance impact here, were talking about whether it works or not. That, and it does work well in some games, as long as you are able to run it:
http://images.nvidia.com/geforce-com/international/comparisons/assassins-creed-syndicate/assassins-creed-syndicate-shadow-quality-interactive-comparison-004-pcss-ultra-vs-high.html
http://images.nvidia.com/geforce-com/international/comparisons/assassins-creed-syndicate/assassins-creed-syndicate-shadow-quality-interactive-comparison-002-pcss-ultra-vs-high.html

Note the increased fidelity and draw distance.

Now compare it to this:
http://www.neogaf.com/forum/showpost.php?p=214763970&postcount=783

If you're going to lose 12-15 fps, it better be worth it.

EDIT: Corrected first link. It was a duplicate of the second one.

what those screens dont show is how pcss breaks msaa coverage

Oh yeah, so much better: https://www.youtube.com/watch?v=QCCdMaLerMg
Or is this a bad video as well?

that video only shows the first 2 cascade levels. if you have gta v its quite easy to see when driving around how awful pcss functions

 

[jorimt]

what those screens dont show is how pcss breaks msaa coverage

You're not going to give up, that much I can see, so I'm going to quit here. "Beware the argument that shuts you up, but fails to convince you."

 

[Collateral22]

After 4 hours in, no black screens nor crashes so far.

Loading time is indeed long for me as well, but I guess it has something to do with my 6 year old HDD & CPU.

It's weird it will just black screen then either hang there where I need to hold down the power button or just say it encountered a program and close.

 

[Tohsaka]

The further I get into this game, the more I'm impressed with it, graphically and gameplay wise. TAA makes 1440p look buttery smooth.

Unfortunately, right after I got rid of the "Hold B to ****" text, now I've got another way more obnoxious message stuck on my screen. The Big Red "Press Y to Disable, Press A to Select" text from after the

Spoiler

bookstore

mission. Can't figure out how to get rid of this one. Might have to run around until I can find 2 praxis kits and experiment with the augs screen. -__-.

This just happened to me, super annoying. Tried relaunching the game and everything, won't go away.

 

[Allnamestakenlol]

After 4 hours in, no black screens nor crashes so far.

Loading time is indeed long for me as well, but I guess it has something to do with my 6 year old HDD & CPU.

Load times are long for me too (1-2 mins), but I've got an ssd. =/

 

[VisceralBowl]

86 save limit?

I can't save anymore.

 

[Peterthumpa]

Is everyone getting weird black vertical bars at each side in FMVs and some menus as well?

 

[The Dear Leader]

86 save limit?

I can't save anymore.

Surely there's a handful of saves you can overwrite of the 86 you have.

 

[VisceralBowl]

Surely there's a handful of saves you can overwrite of the 86 you have.

Yes, and I made a bunch of room. Just a weird number to choose.

 

[·feist·]

Have there been benchmarks and/or reports on vram usage for each texture settings?

GameGPU - bottom of the review page before comments [Russian]

ComputerBase - 2nd page (towards the bottom) [German]

LegitReviews - 3rd page [English]



JaseC:

LegitReviews —— Deus Ex: Mankind Divided DX11 Video Card Benchmarks
http://www.legitreviews.com/deus-ex-mankind-divided-dx11-video-card-benchmarks_185666


Test System:
http://www.legitreviews.com/deus-ex-mankind-divided-dx11-video-card-benchmarks_185666/2

 

[Unknown Soldier]

Is everyone getting weird black vertical bars at each side in FMVs and some menus as well?

Yes. All pre-rendered CGI videos are shown slightly pillarboxed for me, running in 4K resolution (3840x2160). It's weird but at least it's not crashing on CGI playback like some people are experiencing.

 

[texhnolyze]

It's weird it will just black screen then either hang there where I need to hold down the power button or just say it encountered a program and close.

Never heard of that kind of issue. Perhaps you can try switching between fullscreen and exclusive fullscreen?

Load times are long for me too (1-2 mins), but I've got an ssd. =/

I see. Initial loading is the longest one for me. As for loading on the train, it's much shorter. Loading a save is also pretty short, like 5 seconds at most.

I don't remember there's any other loading screens besides those.

GameGPU - bottom of the review page before comments [Russian]

ComputerBase - 2nd page (towards the bottom) [German]

LegitReviews - 3rd page [English]



JaseC:

LegitReviews —— Deus Ex: Mankind Divided DX11 Video Card Benchmarks
http://www.legitreviews.com/deus-ex-mankind-divided-dx11-video-card-benchmarks_185666


Test System:
http://www.legitreviews.com/deus-ex-mankind-divided-dx11-video-card-benchmarks_185666/2

Woah, thanks a bunch!

I wonder if I should try using high textures for my 2gb card. For 1080p, it reaches the maximum vram usage.

I shouldn't, I guess? It'll stutters in occasions.

 

[flipswitch]

Those benchmarks screens from the above are testing in high settings but not ultra?

 

[Unknown Soldier]

Doom was faster on NV for two months after release and it is a known fact that it's Vulkan renderer is solely GCN optimized. Not a valid comparison point until id will optimize the renderer for NV h/w as well.

WarhammerTW isn't faster on 480 than on 1060 even in AMD's half-baked DX12 effort and 1060 is even faster than that in DX11 (no quality difference). False claim.

FM6A isn't really faster either as it depends on the benchmark - I've seen both cards being faster in different benchmarks. So it's a tie basically.

The only "trend" here to speak of is the trend of AMD GE titles being completely unoptimized for NV h/w. For each GE title there are a dozen of non-GE titles where 1060 usually wins.

DX:MD's "Dawn Engine" is a modified version of the Hitman (2016) engine. It was never going to perform well on Nvidia hardware and the benchmarks bear this out with a Fury X beating a 980 Ti and 1070.

I mean if Squenix feels like using an engine purposely de-optimized for the hardware vendor which has 80% of the market then more power to them. I got my copy of the DX:MD for $40 for CDkeys.com and as long as Squenix purposely releases games which are de-optimized for my hardware I'll keep not paying retail price for them. I'm just sad I didn't get in when CDkeys.com had it for $32 a few weeks before release.

 

[·feist·]

Woah, thanks a bunch!

I wonder if I should try using high textures for my 2gb card. For 1080p, it reaches the maximum vram usage.

I shouldn't, I guess? It'll stutters in occasions.

NP. As for memory I've seen multiple sources recommend medium for 2GB cards. YMMV.

Those benchmarks screens from the above are testing in high settings but not ultra?

Do you mean the most recent one I posted from LegitReviews? They tested the GTX 1080 at Ultra/Very High/High for their memory usage findings, though they didn't replicate that for the other cards. Their results have since been updated to include the GTX 970, which was initially omitted. It's *possible* they may further update the benchmarks later.

Other posted reviews have Ultra, mixed and varying settings.

These are not in the OP, so you may or may not have seen them:
https://www.computerbase.de/2016-08/deus-ex-mankind-divided-benchmark/2/
https://www.guru3d.com/articles-pag...-graphics-performance-benchmark-review,1.html

 

[CyberPunked]

There appears to be an issue with the shadow quality setting.
Shadow Quality Very High:

Shadow Quality High:

 

[Allnamestakenlol]

There appears to be an issue with the shadow quality setting.

The hell is going on with this game's graphics settings??

 

[jorimt]

There appears to be an issue with the shadow quality setting.

Ho, add another one to the list. I did notice something was up with High vs. Very High Shadows when testing other settings for quality differences, but I didn't want to sound overbearing. It's sad that the game looks better at lower quality settings.

 

[leng jai]

This port keeps getting worse.

 

[BoosterDuck]

This port keeps getting worse.

woah calm down Lando

at least Nixxes is communicating with us and already sent out a small bugfix this morning

 

[loganclaws]

Most of you probably don't remember, but Human Revolution had a bunch of performance problems when it was released including tons of hitching and long loading times tied to framerate. Nixxes fixed all these issues within about 2 weeks of the release date.

 

[Megasoum]

So on my 1080 if I put everything maxed out in 1080p I get around 20-30fps...

Obviously, the lower I drop the MSAA the higher the FPS goes... I have to disable MSAA to get mostly stable 60fps but even then I have drops from time to time during cutscenes.

 

[Unknown Soldier]

This port keeps getting worse.

I have altered the port graphics. Pray I don't alter them any further.

 

[Azzanadra]

Okay, so a few hours into the game, after I enter that second part of Prague I am hitting 20-30 fps even on LOW with a GTX 970, with massive stuttering. In fact, the performance from low to high is the same. Could my cpu be the problem? Its an AMD Phenom XII 955. Really old, I know. I never bothered to upgrade it because I thought my experience with The Witcher 3 meant the CPU would be okay for at least half a year.

What a shit fucking port. I could max out The Witcher 3 without hairworks at 30-40 FPS, I could straight up max MGSV at 60 fps, and both of those games look better to boot. Heck, The Last of Us on PS3 looked better in some parts.

 

[PFD]

DX:MD's "Dawn Engine" is a modified version of the Hitman (2016) engine. It was never going to perform well on Nvidia hardware and the benchmarks bear this out with a Fury X beating a 980 Ti and 1070.

I mean if Squenix feels like using an engine purposely de-optimized for the hardware vendor which has 80% of the market then more power to them. I got my copy of the DX:MD for $40 for CDkeys.com and as long as Squenix purposely releases games which are de-optimized for my hardware I'll keep not paying retail price for them. I'm just sad I didn't get in when CDkeys.com had it for $32 a few weeks before release.

The Fury X was a direct competitor to the 980 Ti, so it's not out of this world to see it beating the 980 Ti.

 

[Corpekata]

Okay, so a few hours into the game, after I enter that second part of Prague I am hitting 20-30 fps even on LOW with a GTX 970, with massive stuttering. In fact, the performance from low to high is the same. Could my cpu be the problem? Its an AMD Phenom XII 955. Really old, I know. I never bothered to upgrade it because I thought my experience with The Witcher 3 meant the CPU would be okay for at least half a year.

What a shit fucking port. I could max out The Witcher 3 without hairworks at 30-40 FPS, I could straight up max MGSV at 60 fps, and both of those games look better to boot. Heck, The Last of Us on PS3 looked better in some parts.

Of course it's your CPU. It doesn't even match the minimum. The thing's nearly a decade old.

But yes, if the performance stays the same despite changes from low to high, it means your CPU is bottlenecking it.