• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

[PCGAMESHARDWARE.DE] DOOM Benchmarks

Page 5 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Amds opengl drivers on windows perform just fine. The linux side needs performance improvements for sure but the problem is opengl itself.

Besides, i can't be the only 1 to find 5gb oddly specific?

Do they? What windows games do people even use to benchmark opengl performance now? I'd be hard pressed to name any others.

5GB is about (exactly?) the available memory to games on both the PS4 and XB1, so...probably not a coincidence there.
 
Amds opengl drivers on windows perform just fine. The linux side needs performance improvements for sure but the problem is opengl itself.

Besides, i can't be the only 1 to find 5gb oddly specific?

I think Doom 2016 is the most complex OpenGL game by far, and AMD is usually behind Nvidia in OGL even on more basic stuff

80123.png



pretty clear who is losing more with OpenGL...



didnt they pulled the exact same feat with with the unpacked "Textures" on doom 3?
yeap they did..

Doom 3 ultra required or recommended 512MB, but pretty much all high end cards had 256MB and mainstream was 128MB.
 
Last edited:
http://www.overclock.net/t/1556226/gamegpu-wolfenstein-the-old-blood-benchmark

Last idtech game.
Opengl worked just fine and was performant.
Nice spin though

well although this game is based on id tech 5 the last game made by id was RAGE, which was a disaster specially on AMD cards at launch (stability problems), now, Wolfenstein Old Blood, correct me if I'm wrong but it runs on OGL 3.2 or something like that (close to DX9), while Doom uses a new engine and targets OGL 4.5

also, the results from Talos are valid, Nvidia lost a lot less performance with OGL, the many Linux benchmarks are also a good indication, massive advantage for Nvidia on more complex games.

and there articles from open source developers like this
https://dolphin-emu.org/blog/2013/09/26/dolphin-emulator-and-opengl-drivers-hall-fameshame/
 
On the CPU optimization side, turning on HT on the $1000 Intel CPU halves the FPS. Nice bug/optimization there.

doom_proz.jpg

Whatever is going on there isn't affecting all HW-E and/or 5960Xs. Mine is working as intended with HT enabled, meanwhile Carfax83's system appears to be plagued by it over in the PC Gaming thread.
 
Well done Abwx, nice way to prove my point. That tweet from AMD engineer saying they're working on it kinda supports my narrative too, everything looks good from AMD on paper.
 
Well done Abwx, nice way to prove my point. That tweet from AMD engineer saying they're working on it kinda supports my narrative too, everything looks good from AMD on paper.

Or maybe the dev worked a bit more closely with Nvidia on the final game than AMD so AMD had less chance to optimise it on Ultra settings? The Beta results on lower settings worked fine on AMD cards.

Or do you think it is a coincidence that Nvidia used Doom to show off Vulkan at the GTX1080 launch and not the Talos Principle??
 
http://www.overclock.net/t/1556226/gamegpu-wolfenstein-the-old-blood-benchmark

Last idtech game.
Opengl worked just fine and was performant.
Nice spin though

That's the problem. Anytime benchmarks come out, people start spinning like crazy. Oh, this means that Nvidia has this performance, or this means that.

Average it over a long period of games instead and you find the following:

Nvidia has better launch performance
AMD has better long term performance
Nvidia has better launch performance at games
Nvidia has more games that allow SLI
AMD has better CF scaling
AMD can have better performance long term if they release drivers for a particular game and it's not ignored.
etc.

Instead, people quote a single instance of something and extrapolate it as truth to all instances, just to support their favorite brand. Nvidia and AMD both have pros and cons, but it's annoying to see the spin doctors on this forum try to make it seem like one GPU solution is the best for everyone.
 
Hey guys, on topic plz.

I think the game needs better optimization, here is hoping for SLI/CF support with additional patches. It played pretty well though on my system on medium-high if I do remember, that is only using one 290.
 
I remember AMDs graphics cards having problems to catch Nvidias in Rage and recent Wolfenstein games.

Not that the first implementations were smooth or amd didnt have deiver bugs. I am sure there were but to keep quoting misinformation does a disservice to us all.

Yep, I remember having to manually tweak my cfg files for Rage to get acceptable texture streaming. It was horrendous at launch. I had a 5870 2GB and the GTX had a 460 1GB, using the auto settings, game look awful on the Radeon but her 460 seemed to stream the textures a little better.

Man was that game a cluster-screw on AMD hardware.
 
It's using 4.3 with extensions from 4.4 and 4.5. There is no advantage in initializing a 4.5 context.

We use as much or little of 4.5 on NVIDIA as we do on AMD. No difference.

https://twitter.com/axelgneiting/status/731958687472877569
https://twitter.com/axelgneiting/status/731573399172960256

its strange 4.3 is being reported for AMD while 4.5 for Nvidia. We likely won't know the details of what they did. AMD will just be stuck trying to fix whatever it was. I suspect the vulkan update will be messed up as well. Somehow they will manage to mess it up for AMD users.

To me there is no doubt they could have fixed this before launch. They wrote the code for the game and should have tested it. It seems a trend that their games have issues with AMD while even smaller developers don't.
 
https://twitter.com/axelgneiting/status/731958687472877569
https://twitter.com/axelgneiting/status/731573399172960256

its strange 4.3 is being reported for AMD while 4.5 for Nvidia. We likely won't know the details of what they did. AMD will just be stuck trying to fix whatever it was. I suspect the vulkan update will be messed up as well. Somehow they will manage to mess it up for AMD users.

To me there is no doubt they could have fixed this before launch. They wrote the code for the game and should have tested it. It seems a trend that their games have issues with AMD while even smaller developers don't.

Forgive me if I misunderstood but it was running fine on AMD and then they spent time "optimizing" with nVidia and now it runs like crap on AMD. If that's it should anyone really be surprised?
 
Forgive me if I misunderstood but it was running fine on AMD and then they spent time "optimizing" with nVidia and now it runs like crap on AMD. If that's it should anyone really be surprised?

No, the beta was running medium settings. The AMD cards seem to have a problem with Ultra settings, which were not in the beta.

Relax, I am sure AMD will put out a new driver soon and it'll run fine. It's not a big deal.
 
No, the beta was running medium settings. The AMD cards seem to have a problem with Ultra settings, which were not in the beta.

Relax, I am sure AMD will put out a new driver soon and it'll run fine. It's not a big deal.

I don't need to relax. Me getting up tight wouldn't effect anything and I've learned not to bother.

Thanks for the clarification. So on Medium settings AMD's performance is still OK?
 
Back
Top