People were interested in seeing some AMD benchmarks for the just released Dirt Showdown and I'm happy to oblige. Here are some comparative results for both of my main GPUs.
Remember, based on theoretical numbers these two GPUs are about as powerful. The R7 of course has more VRAM than the 760 (2GB vs 4GB) but that rarely gets utilized. So, without further ado, let's benchmark this!
So here are the Nvidia numbers. As you can see, the game performs very well and only goes below the magical 60 FPS mark on Ultra and even there it's at a reasonable 52 FPS. Note that these are average framerates and the game can occasionally drop about 20 FPS for a moment. The performance drop-off is fairly small up until Ultra, where the performance drops quite radically.
And now we come to the newbie of my hardware, the R7 370 which I have been using for the past two weeks. While Dirt Showdown is apparently playable on the RadeonSI driver, provided that you compile it yourself, I stuck to the Catalyst 15.7 driver for this benchmark. Mainly because I'm extremely lazy and won't compile my own GPU drivers. It's worth noting that the R7 runs a lot slower on Ultra Low, it's only about half the performance of the 760. However, the performance also seems to drop a lot slower. And when reaching Ultra the R7 actually catches up to the 760.
Here you can see both results on the same chart for easier reading:
So generally the R7 performs quite a bit worse than the 760 initially. I haven't quite figured out what caused the sudden drop in performance for the 760 on Ultra. It could be the VRAM but somehow I don't think that's the case. Regardless, the game was completely playable on both cards and I congratulate Virtual Programming for a fine performing port. I haven't actually played the game yet, I simply ran these benchmarks but if the benchmarks give any idea about how well the game is going to run, then this will probably be a pleasant game to play framerate-wise.
If you want to see additional Nvidia numbers and a little bit of gameplay, you can head on over to Liam's preview article where he tested the game with a GTX 560 Ti and a GTX 970.
I'll reserve a spot for Dirt Showdown for the Friday Livestream so you can see my first impressions of the game there. Article taken from GamingOnLinux.com.
Remember, based on theoretical numbers these two GPUs are about as powerful. The R7 of course has more VRAM than the 760 (2GB vs 4GB) but that rarely gets utilized. So, without further ado, let's benchmark this!
So here are the Nvidia numbers. As you can see, the game performs very well and only goes below the magical 60 FPS mark on Ultra and even there it's at a reasonable 52 FPS. Note that these are average framerates and the game can occasionally drop about 20 FPS for a moment. The performance drop-off is fairly small up until Ultra, where the performance drops quite radically.
And now we come to the newbie of my hardware, the R7 370 which I have been using for the past two weeks. While Dirt Showdown is apparently playable on the RadeonSI driver, provided that you compile it yourself, I stuck to the Catalyst 15.7 driver for this benchmark. Mainly because I'm extremely lazy and won't compile my own GPU drivers. It's worth noting that the R7 runs a lot slower on Ultra Low, it's only about half the performance of the 760. However, the performance also seems to drop a lot slower. And when reaching Ultra the R7 actually catches up to the 760.
Here you can see both results on the same chart for easier reading:
So generally the R7 performs quite a bit worse than the 760 initially. I haven't quite figured out what caused the sudden drop in performance for the 760 on Ultra. It could be the VRAM but somehow I don't think that's the case. Regardless, the game was completely playable on both cards and I congratulate Virtual Programming for a fine performing port. I haven't actually played the game yet, I simply ran these benchmarks but if the benchmarks give any idea about how well the game is going to run, then this will probably be a pleasant game to play framerate-wise.
If you want to see additional Nvidia numbers and a little bit of gameplay, you can head on over to Liam's preview article where he tested the game with a GTX 560 Ti and a GTX 970.
I'll reserve a spot for Dirt Showdown for the Friday Livestream so you can see my first impressions of the game there. Article taken from GamingOnLinux.com.
0 Likes
Some you may have missed, popular articles from the last month:
The comments on this article are closed.
15 comments
What settings does the game allow you to change manually? If possible, you could try reducing the texture quality while leaving the rest on ultra. That should reduce the required VRAM, so might improve the performance on GTX 760 if VRAM is the issue.
0 Likes
ElectricPrism Aug 18, 2015
Phoenix put out a article on this too
nVidia vs AMD Linux Gaming DIRT
Thanks for sharing the data :)
nVidia vs AMD Linux Gaming DIRT
Thanks for sharing the data :)
2 Likes, Who?
Quoting: cdnr1why aren't you using the latest nvidia driver352.30 is the latest stable driver. 355 is a beta driver.
0 Likes
Awesome job Samsai, definitely looking forward to more of these articles in the future where possible :)
0 Likes
there must be some sort of obvious driver bottleneck here on AMD. The fact that it catches up at ultra and that overall the performance is holding quite well.
fAilMD again. Keep hearing about the opensource drivers catching up to catalyst, then we remember catalyst is a lot slower than it should be. How are AMD going to manage to get all existing openGL games ( 1400 ) upto scratch and ready for SteamOS ? Vulcan may help new titles but Vulcan is only going to be supported on the latest cards and that leaves all 1400 ( maybe near 2000 ) by the time they get the performance right, games in a weak or broken position.
honestly, Valve should of made SteamOS exactly as they have done now, but with a fixed performance baseline using a 780Ti/980Ti and allowed for AMD as a option just because its possible. Then each year support better and faster Nvidia cards.
Which is sort of happening but they should just come out and state it as a target. It might at the very least get AMD to wake up.
fAilMD again. Keep hearing about the opensource drivers catching up to catalyst, then we remember catalyst is a lot slower than it should be. How are AMD going to manage to get all existing openGL games ( 1400 ) upto scratch and ready for SteamOS ? Vulcan may help new titles but Vulcan is only going to be supported on the latest cards and that leaves all 1400 ( maybe near 2000 ) by the time they get the performance right, games in a weak or broken position.
honestly, Valve should of made SteamOS exactly as they have done now, but with a fixed performance baseline using a 780Ti/980Ti and allowed for AMD as a option just because its possible. Then each year support better and faster Nvidia cards.
Which is sort of happening but they should just come out and state it as a target. It might at the very least get AMD to wake up.
1 Likes, Who?
Quoting: mr-eggthere must be some sort of obvious driver bottleneck here on AMD. The fact that it catches up at ultra and that overall the performance is holding quite well.
fAilMD again. Keep hearing about the opensource drivers catching up to catalyst, then we remember catalyst is a lot slower than it should be. How are AMD going to manage to get all existing openGL games ( 1400 ) upto scratch and ready for SteamOS ? Vulcan may help new titles but Vulcan is only going to be supported on the latest cards and that leaves all 1400 ( maybe near 2000 ) by the time they get the performance right, games in a weak or broken position.
honestly, Valve should of made SteamOS exactly as they have done now, but with a fixed performance baseline using a 780Ti/980Ti and allowed for AMD as a option just because its possible. Then each year support better and faster Nvidia cards.
Which is sort of happening but they should just come out and state it as a target. It might at the very least get AMD to wake up.
being equal on Ultra would probably mean something completely different. from my experience this would be limit at which rest of the computer can feed to gpu, but that would only be possible to be sure about if benchmark included cpu utilization.
when you hit cpu bottleneck it doesn't matter if you have Titan X or Iris 4000. computer simply cannot feed frames which could be rendered on gpu and you end up with under utilized gpu where both will render same amount of frames as long as they can since limitation is elsewhere
also, there is one shiny spot in future. Vulkan and its simpler drivers should be able to create equal standings for all vendors. AMD is probably betting on the fact that next gen could finally enable them to catch up with times and get rid of ATi past with terrible drivers
Last edited by vulture on 19 August 2015 at 2:15 pm UTC
0 Likes
DamonLinuxPL Aug 19, 2015
Hmm. If I good understand new eON games have still disabled some OPenGL extension like ARB_texture_storage on Catalyst driver and this causes worst performance. So we can run this like the witcher2 with command line with command "--eon_disable_catalyst_workarounds"
And now check performamance. Anyone can test it on Catalyst 15.7?
And now check performamance. Anyone can test it on Catalyst 15.7?
1 Likes, Who?
The sad part is that in Linux that 760 would probably outperform my 290x crossfire...
0 Likes
LinuxGamesTV Aug 20, 2015
Quoting: mr-eggthere must be some sort of obvious driver bottleneck here on AMD. The fact that it catches up at ultra and that overall the performance is holding quite well.
fAilMD again. Keep hearing about the opensource drivers catching up to catalyst, then we remember catalyst is a lot slower than it should be. How are AMD going to manage to get all existing openGL games ( 1400 ) upto scratch and ready for SteamOS ? Vulcan may help new titles but Vulcan is only going to be supported on the latest cards and that leaves all 1400 ( maybe near 2000 ) by the time they get the performance right, games in a weak or broken position.
honestly, Valve should of made SteamOS exactly as they have done now, but with a fixed performance baseline using a 780Ti/980Ti and allowed for AMD as a option just because its possible. Then each year support better and faster Nvidia cards.
Which is sort of happening but they should just come out and state it as a target. It might at the very least get AMD to wake up.
Again, AMDGPU is the next-gen AMD driver (FLOSS and BLOB) in Kernel 4.2.x.
If you don't have a rolling realese from your Distribution, you must update self the Kernel or you don't benefit from it.
You all cry about AMD driver is bad, etc. Yes they are bad but AMD work hand in hand with the FLOSS driver FOR AMD Cards Dev's to make a better BLOB driver from the scratch.
The kernel modules AMDGPU will be the nerve center from the new FLOSS and BLOB driver for AMD GPU's.
For the FLOSS driver brings MESA the openGL stuff and for the BLOB driver, the BLOB but not only the openGL stuff. The BLOB brings the reclocking stuff, too. The FLOSS should don't have the reclocking stuff.
And than comes Vulkan, not Vulcan.
Yes the new BlOB drivers from AMD and NVIDIA have both: openGL and Vulkan.
0 Likes
Quoting: BdMdesigNYou all cry about AMD driver is bad, etc. Yes they are bad but AMD work hand in hand with the FLOSS driver FOR AMD Cards Dev's to make a better BLOB driver from the scratch.
And than comes Vulkan, not Vulcan.
Yes the new BlOB drivers from AMD and NVIDIA have both: openGL and Vulkan.
people aren’t 'crying'
I stayed with my AMD GPU for two years after leaving windows with poor performance, after about 6 months I decided to not dual boot. So that was a year and a half of terrible gaming and desktop experiences. In the end I had no choice, it had to be Nvidia.
4 years on its the same.
My point, staying on topic was merely that a lot of games have driver specific tweaks because the devs are often too rushed or inexperienced enough to get things working right, It should not have to be upto AMD or Nvidia to patch companies games endlessly for bugs and make work around's in their driver ... but they do. Thats a reality.
Given that fact,my point was how are AMD going to manage patching existing / new AAA SteamOS/Linux games when they haven’t had a decent driver let alone game / software patches ? the answer is AMD will ALWAYS perform worse in those situations but will be masked with newer more powerful hardware to some extent.
its a genuine question.. how will any AMD card like for like against Nvidia even when the drivers become better manage to support the existing Patches that are required? And given they are always weeks if not months late with game patches vs Nvidia on Windows, how will they manage to keep up in this overly congested Gaming market when they also have two consoles to manage ( of course that driver situation is slightly different but even so ) ?
AMD are almost out of the game.
FYI your vulkan ( not vulcan my sp bad ) point is not really going to address:
a.) 1400 - 1600 games that already exist that require workarounds in drivers OR just plain good old OpenGL performance ( how long do we have to wait ? a year ? how many steam machine will have nvidia hardware in ? )
b.) Unity's work on improving their in house OpenGL code base and only marginal interest in Vulkan. So, the biggest engine right now for pushing out crossplatform/Linux games, meaning a lot of new games on SteamOS will still use OpenGL on AMD
c.) Vulkan actually being adopted whole sale against OpenGL AND requiring lots of developer investment even though the API is cleaner. There have been claims that just like DX12, unless you were really gunning for massive performance optimisations and required large scale multi threading on your title you wouldnt see huge Vulkan gains. But lets wait and see.
TLDR
1. AMD have got better
2. Their open source approach is commendable.
3. There still kind of sucky today and today is a day I want to play GOL
Last edited by on 20 August 2015 at 1:57 pm UTC
1 Likes, Who?
Quoting: BdMdesigNFor the FLOSS driver brings MESA the openGL stuff and for the BLOB driver, the BLOB but not only the openGL stuff. The BLOB brings the reclocking stuff, too. The FLOSS should don't have the reclocking stuff.
And than comes Vulkan, not Vulcan.
Yes the new BlOB drivers from AMD and NVIDIA have both: openGL and Vulkan.
Actually the reclocking support will come in at the kernel level. So, the AMDGPU kernel driver will provide the power management support for both the open source RadeonSI driver and Catalyst once everything falls into place. I predict that AMD will kill off proprietary drivers on Linux within 12 months, RadeonSI is now at OpenGL 4.1 and almost OpenGL 4.2. Another 12 months of development and we should have a very good and fast open source driver.
1 Likes, Who?
Quoting: mr-eggc.) Vulkan actually being adopted whole sale against OpenGL AND requiring lots of developer investment even though the API is cleaner. There have been claims that just like DX12, unless you were really gunning for massive performance optimisations and required large scale multi threading on your title you wouldnt see huge Vulkan gains. But lets wait and see.
that is so... crap
adoption of Vulkan on linux has much larger benefits than DX12 has on windows. performance being least important of all. OpenGL has whole lot of other problems that DX before 12 didn't suffer from
- different core profiles supported by vendors (Vulkan solves that)
- shaders being different per vendor (SPIR-V does not allow that)
- differences between implementations in performance or functionality between vendors (Vulkan will ship with enforced conformance tests)
- different implementations per mobile/desktop (Vulkan again solves this)
on Linux, all the problems of OpenGL will be fixed with Vulkan, on Windows DX was in much better state as far as problems like these go.
if you only see performance gains... stop looking. you are blind
just implementing OpenGL on top of Vulkan would solve a lot more even if it had 50% performance drop. new faster hardware is coming a lot faster than fixing of inherent problems
Last edited by vulture on 20 August 2015 at 9:46 pm UTC
0 Likes
Quoting: vulturethat is so... crap, you are blind
~
You put your point across so eloquently. Next time instead of surfing words and taking things out of context why don’t you read and then relay clearly, with your impressive knowledge of the English language.
here..
Quoting: mr-eggc.) Vulkan actually being adopted whole sale against OpenGL AND requiring lots of developer investment even though the API is cleaner. There have been claims that just like DX12, unless you were really gunning for massive performance optimisations and required large scale multi threading on your title you wouldnt see huge Vulkan gains. But lets wait and see.
see I qualified my statement. It was not a definitive.
http://www.saschawillems.de/?p=1886
QuoteIt won’t magically speed up your rendering
Don’t expect your single threaded renderer to be much faster after switching from OpenGL over to Vulkan (unless it’s ancient and using OpenGL immediate mode). Vulkan is all about getting rid of CPU bottlenecks often caused by driver overhead when rendering complex scenes (aka “I can haz more drawcalls plz?”). So if you want the performance gains possible with Vulkan you need to go all in on multi threaded resource management
Is he the most knowledgeable and influential person out there ? No. Does he know more this than you? Probably YES.
Vulkan, is probably not overnight going to fix Linux Gaming. Linux needs the same attention on game development as any other platform does.
IMO Wayland offers more excitement in the desktop space.
but yea. .. crap n stuff ...
ohh btw..
http://www.phoronix.com/scan.php?page=news_item&px=KWIN-No-Vulkan-Plans
Out today ^ Given KDE is supposed to be the most popular desktop environment. Not everyone is throwing the kitchen sink into vulkan. So your dream non GL desktop might have to wait a good bit longer.
Last edited by on 21 August 2015 at 5:16 pm UTC
0 Likes
Patreon
PayPal
See more from me