DOOM (2016) first benchmark review

This is copy pasta from another forum, but I find it very interesting that this guy with a 390x says DOOM pushes his GPU to 76c... I found the opposite, GPU was running eerily cool, even compared to the beta
sean8102;5273006 said:
Yeah I saw some people saying that about the beta as well. I played the beta but never checked what version of OpenGL it was using.

Also this game seems to push my GPU harder than any other game I've played so far. I've seen my GPU peak to 76 degrees Celsius. Highest I've seen before was about 72 or 73. I got a new case recently and took a good long time cleaning everything and my cable management is really good so I guess it just stresses the card a lot!

Edit: Found a post on twitter from a ID dev saying that AMD is working on improving performance.

Click to expand...
Click to expand...
 
This looks about accurate to me - running a stock GTX970 and dropping the res down to 2560x1440 for smooth gameplay I'm getting very similar frame rates to those shown in the graph, at least from the first couple of hours into the game.
 
piratepress said:
I agree that it runs fantastic. I'm running it at Ultra 1080 on my 970 and it never drops below 60 FPS. I could probably bump it up to 1440 and still gets 40 FPS.
Click to expand...

This is the part where I'd love to tell all you DX heathens to kneel before the amazingness that is OpenGL.

But looking at the screenshots it appears the devs are only using a 256-color palette so... ¯\_(ツ)_/¯
 
jwcalla said:
This is the part where I'd love to tell all you DX heathens to kneel before the amazingness that is OpenGL.

But looking at the screenshots it appears the devs are only using a 256-color palette so... ¯\_(ツ)_/¯
Click to expand...
lol i agree....Maybe its just the outside areas are done that way? I think dx9 games looked almost as good as what were seeing these days (sometimes)
 
First benchmark done, will update with other results ;)

I'm running the benchmark in the first room after getting the shotgun, like the PCGH review I run around this room.

Never mind, can't compare to the review in OP because they had killed all the demons, I ran the benchmark in the same room doing a similar run but the demons were still alive, and I killed the one at the door every time

1440p MAX settings (nightmare) FOV 105 - 980Ti@ 1480-8000
2016-05-14 23:56:24 - DOOMx64
Frames: 3208 - Time: 30797ms - Avg: 104.166 - Min: 87 - Max: 134

1440p MAX settings (shadows ultra) FOV 105 - 980Ti@ 1480-8000
2016-05-14 23:37:41 - DOOMx64
Frames: 3455 - Time: 32578ms - Avg: 106.059 - Min: 87 - Max: 128

1080p MAX settings (nightmare) FOV 105 - 980Ti@ 1480-8000
2016-05-14 23:47:16 - DOOMx64
Frames: 4581 - Time: 31922ms - Avg: 143.506 - Min: 118 - Max: 171

1080p MAX settings (shadows ultra) FOV 105 - 980Ti@ 1480-8000
2016-05-14 23:45:46 - DOOMx64
Frames: 4124 - Time: 27063ms - Avg: 152.385 - Min: 135 - Max: 195


This is a different benchmark run, I'm running around the weird demon altar thing in the same room three times

1440p MAX settings (nightmare) FOV 105 - 980Ti@ 1480-8000
2016-05-15 00:01:36 - DOOMx64
Frames: 1833 - Time: 20266ms - Avg: 90.447 - Min: 86 - Max: 95

1440p MAX settings (shadows ultra) FOV 105 - 980Ti@ 1480-8000
2016-05-15 00:02:10 - DOOMx64
Frames: 2547 - Time: 26562ms - Avg: 95.889 - Min: 91 - Max: 101
 
Last edited:
Any opinions on what settings a stock GTX 950 would run? Summer is coming so I probably won't have time to play anything until winter again but.. Just curious. 960 seems to fair well, but... 2 gig of RAM on the 950 has me wondering how much I'd be turning things down.
 
GameGPU has the GTX950 showing 1920x1080 'High' settings (not Ultra) at a minimum fps of 40, average 47.
 
Is about the same score I get at the settings I run it at, and that seems fairly smooth to me. You should be OK at that setting I think - the game seems to need around 3GB VRAM for ultra settings, but there's no graphic for 'high' so I'm not sure how much that demand lessens. Presumably a fair bit, as there are lots of 2GB cards that seem fairly proportional in performance at that level without sudden dips in the minimum fps figures.
 
I cannot speak to this review, but I can speak to my own testing. Fury X is just a few percentage behind 980 Ti at 4K using latest drivers. The game is playable at the same settings between Fury X and GTX 980 Ti.
 
Brent_Justice said:
I cannot speak to this review, but I can speak to my own testing. Fury X is just a few percentage behind 980 Ti at 4K using latest drivers. The game is playable at the same settings between Fury X and GTX 980 Ti.
Click to expand...

A reference 980ti?

Can you please specify what clocks it effectively boosts at in your review, that's something I really like about PCGH charts
 
Ieldra said:
A reference 980ti?

Can you please specify what clocks it effectively boosts at in your review, that's something I really like about PCGH charts
Click to expand...
I think a publication has to either go all reference models or like pcgameshardware all AIB.
Thankfully we are lucky to be able to use both sets of data but from different places as both definitely have their uses, but we do see fans who argue on some of those other sites using either statistic usually out of context :)
Cheers
 
Brent_Justice said:
I cannot speak to this review, but I can speak to my own testing. Fury X is just a few percentage behind 980 Ti at 4K using latest drivers. The game is playable at the same settings between Fury X and GTX 980 Ti.
Click to expand...
Is that with Ultra or Nightmare settings Brent?
I assume Ultra as the Fury X is only 4GB.
I take it the figures spread a bit when looking at 1440p?
Thanks
 
CSI_PC said:
I think a publication has to either go all reference models or like pcgameshardware all AIB.
Thankfully we are lucky to be able to use both sets of data but from different places as both definitely have their uses, but we do see fans who argue on some of those other sites using either statistic usually out of context :)
Cheers
Click to expand...

My main issue is what if the benchmark run is particularly short so the card doesn't heat up much, could it run at say 1240mhz for the duration of the benchmark, but in extended sessions throttle to 1200? My card doesn't go above the boost written on the box without me giving it a clock offset
 
Ieldra said:
Never mind, can't compare to the review in OP because they had killed all the demons, I ran the benchmark in the same room doing a similar run but the demons were still alive, and I killed the one at the door every time

1440p MAX settings (nightmare) FOV 105 - 980Ti@ 1480-8000
2016-05-14 23:56:24 - DOOMx64
Frames: 3208 - Time: 30797ms - Avg: 104.166 - Min: 87 - Max: 134

1440p MAX settings (shadows ultra) FOV 105 - 980Ti@ 1480-8000
2016-05-14 23:37:41 - DOOMx64
Frames: 3455 - Time: 32578ms - Avg: 106.059 - Min: 87 - Max: 128

This is a different benchmark run, I'm running around the weird demon altar thing in the same room three times

1440p MAX settings (nightmare) FOV 105 - 980Ti@ 1480-8000
2016-05-15 00:01:36 - DOOMx64
Frames: 1833 - Time: 20266ms - Avg: 90.447 - Min: 86 - Max: 95

1440p MAX settings (shadows ultra) FOV 105 - 980Ti@ 1480-8000
2016-05-15 00:02:10 - DOOMx64
Frames: 2547 - Time: 26562ms - Avg: 95.889 - Min: 91 - Max: 101
Click to expand...

Looking at those figure, depending upon the area has a moderate influence on performance.
Although amusing how the minimum for the bottom Ultra is actually better than the top test that is faster on average and max, including same context Nightmare for top also beating that bottom Ultra :)
Cheers
 
Ieldra said:
My main issue is what if the benchmark run is particularly short so the card doesn't heat up much, could it run at say 1240mhz for the duration of the benchmark, but in extended sessions throttle to 1200? My card doesn't go above the boost written on the box without me giving it a clock offset
Click to expand...
Yeah good point that was an issue in some games for the reference 980ti, according to some review sites.
Not sure about AMD's recent reference cards cooling-throttling behaviour if stressed in a game for a moderate length of time.
Cheers
 
CSI_PC said:
Looking at those figure, depending upon the area has a moderate influence on performance.
Although amusing how the minimum for the bottom Ultra is actually better than the top test that is faster on average and max :)
Cheers
Click to expand...

yeah I was surprised at how small the hit was from Nightmare shadows, it's more pronounced at the center of the room because of the altar thing being a light source and casting more shadows.

yeah the first set of results were just with me running around the perimeter of the room, i think the variation in min/max is due to where i pointed the camera actually, maybe in top run whole FoV was taken up by a wall or something
 
is there a timedemo/benchmark utility built-in? or are you just using fraps?
 
Ieldra said:
yeah I was surprised at how small the hit was from Nightmare shadows, it's more pronounced at the center of the room because of the altar thing being a light source and casting more shadows.

yeah the first set of results were just with me running around the perimeter of the room, i think the variation in min/max is due to where i pointed the camera actually, maybe in top run whole FoV was taken up by a wall or something
Click to expand...
You also noticed in your results though that the 1st test for Nightmare actually performed better than the 2nd test Ultra :)
That is for Average and Maximum fps, that shadow is definitely dependant upon room-area/positioning it seems.
Cheers
 
Reality said:
is there a timedemo/benchmark utility built-in? or are you just using fraps?
Click to expand...
fraps, I was trying to emulate the benchmark run used for the PCGH review but then i noticed they had cleared the room of mobs so useless comparison

CSI_PC said:
You also noticed in your results though that the 1st test for Nightmare actually performed better than the 2nd test Ultra :)
That is for Average and Maximum fps, that shadow is definitely dependant upon room-area/positioning it seems.
Cheers
Click to expand...

The max was higher for Nightmare shadow vs ultra shadow but that's just consequence of me not being a robot and repeating the exact same run each time :p avg was slightly higher for ultra shadow

the weirdest thing is my gpu is relatively cool, haven't seen it break 62c and I forgot doom was open while i was posting the benchmarks, left it running for like 4 hours :p I've seen 390x users complain about very high temps with doom, that is really odd
 
By the way guys, it seems the 1080 was running ultra settings in the demo, not nightmare
 
how much does the acural vram usage go up with nightmare texture and shadows
at 2560x1440 ultra its using about 3g going by gpu-z but it wont let me enable either of them saying i need minimum 5g
it also seems to use very little system ram
 
Ieldra said:
The demonstration, the video of the 1080 running doom with vulkan
Click to expand...
You getting that info from a site?
Damn you may be right, they said the settings were maximum but I am seeing a discrepancy with their Youtube comment saying Ultra.
Cheers
 
Last edited:
OK, so I just got to play a couple hours and I love the game. $60 well spent in my opinion.

As for performance, I set everything to max at first and was in the 40-60 range. Went in and set lights and shadows to High instead of Ultra and now everything is 70-110. VRAM usage was 2.8GB according to Afterburner and GPU usage was a steady 100%. So overall I'm pretty pleased with the performance of my 290X and I'm not too upset I had to turn down 2 settings I wont notice anyway from Ultra to High. Oh and Im still on the old 16.3.2 AMD drivers so maybe I'll see an increase if/when I switch to 16.5.2.
 
CSI_PC said:
Bah you read my post before my edit :)
They said it was max settings but yeah I now notice Ultra written in a couple of places including their youtube channel (where ironically they also say max setting).
Cheers
Click to expand...

I think it's just because ultra is the highest preset, either way it's all meaningless until we get the vulkan version to test
 
jwcalla said:
Actually I shouldn't be too harsh. Doom3 only had one color: black.

So this is an improvement.
Click to expand...

I lold. It's pretty damning that my most vivid memory of Doom 3 was trying to figure out how to use flashlight without holstering gun, and my surprise when I discovered you couldn't. Lol.
 
CSI_PC said:
Is that with Ultra or Nightmare settings Brent?
I assume Ultra as the Fury X is only 4GB.
I take it the figures spread a bit when looking at 1440p?
Thanks
Click to expand...

4K at SMAA and "Ultra" global setting, which is playable on both.

You cannot set Nightmare settings on Fury X, it gives a warning message that a video card with 5GB or greater dedicated VRAM is required.

Also, I do notice a few hiccups on the Fury X while playing, as if it is streaming in textures, which makes sense. This game probably does not use Fury X's dynamic VRAM capability at all.

At 1440p its a wash, and not very challenging, both GPUs are underutilized a great deal at 1440p. I have to use 4K simply for the reason that only at 4K does any performance advantage show itself.
 
I'm using TAA @ Ultra with motion blur disabled (the game moves so fast that motion blur is a terrible idea), It's maintaining mid 70's and above @ 3440x1440. The game has mouse smoothing enabled by default, you need to drop in +set m_smooth 0 into the startup options to kill it off.
 
I hate to bump the same question in both Doom threads, but I must know.....

Anyone tried benching the game on an old school i5-750/760 equivalent?

(I'd be completely content with a mixture of med-high settings)
 
Kor said:
I'm using TAA @ Ultra with motion blur disabled (the game moves so fast that motion blur is a terrible idea), It's maintaining mid 70's and above @ 3440x1440. The game has mouse smoothing enabled by default, you need to drop in +set m_smooth 0 into the startup options to kill it off.
Click to expand...

Ultra too with Nano, turning off motion blur really helps performance I am also in the 70s and 80s with it on low (I assuming that is what you meant) at 3440x1440. Turned off mouse smoothing as well. FOV 110.

I turned down sharpening from 2 to 1.5, aliasing seems to be significantly less and I like it much better.
 
noko said:
Ultra too with Nano, turning off motion blur really helps performance I am also in the 70s and 80s with it on low (I assuming that is what you meant) at 3440x1440. Turned off mouse smoothing as well. FOV 110.

I turned down sharpening from 2 to 1.5, aliasing seems to be significantly less and I like it much better.
Click to expand...

Motion Blur off, not low. You can disable it completely in the Video menu rather than just bumping down the quality.
 
Kor said:
Motion Blur off, not low. You can disable it completely in the Video menu rather than just bumping down the quality.
Click to expand...

Your right, ok it is off. This adds a lot of performance and since game runs so smooth to begin with I also don't see the need for it.
 
dasa said:
how much does the acural vram usage go up with nightmare texture and shadows
at 2560x1440 ultra its using about 3g going by gpu-z but it wont let me enable either of them saying i need minimum 5g
it also seems to use very little system ram
Click to expand...

Low to mid 4GB range usually but that jumped up to 5GB on Kadingir Sanctum
 
You must log in or register to reply here.
Back
Top