Hey there folks!
Today there are two tests to showcase. One is showcasing Dota 2 on a GTX 680 in both Operating Systems, and the other is showcasing Dota 2 on an Intel HD 4600 iGPU in both Operating Systems.
The first video showcases the test running with the GTX 680:
The second video showcases the test running with the Intel HD 4600:
In case you want to skip, and just grab the important details:
I start summarizing the test in the Nvidia video at about 7:20.
I start summarizing the test in the Intel video at about 7:06.
For the Nvidia tests, both systems were running the latest 331 drivers. Ubuntu had 331.13 and Windows had 331.40.
For the Intel tests, Ubuntu was using Mesa 9.3 and Windows was using the 9.18.10.3165. To my knowledge, these were the latest drivers available for both platforms.
With both these tests, we can see the difference of how a discrete graphics solution performs against an integrated solution on Ubuntu or Linux in general. They aren't representative of an entire community of hardware and configurations, but at least it should help provide us with a point of reference.
In the Nvidia tests, Dota 2 was maxed out at 1920x1080. In the Intel tests, Dota 2 was set to lowest possible settings at 1920x1080.
The tests were done on relatively some of the latest hardware available in the market, with the key components being:
CPU - Intel i7-4770 3.4GHz Haswell
iGPU - Intel HD 4600
GPU - Nvidia GTX680 GDDR5 2GB
AMD discrete and integrated graphics users may experience vastly different results. Older Intel integrated solutions may also experience very different results.
As always, please help correct me if I made any mistakes and advice is always welcome. Thank you to all who have given me corrections or advice in the past, they have helped me make better videos.
Today there are two tests to showcase. One is showcasing Dota 2 on a GTX 680 in both Operating Systems, and the other is showcasing Dota 2 on an Intel HD 4600 iGPU in both Operating Systems.
The first video showcases the test running with the GTX 680:
YouTube videos require cookies, you must accept their cookies to view. View cookie preferences.
Direct Link
Direct Link
The second video showcases the test running with the Intel HD 4600:
YouTube videos require cookies, you must accept their cookies to view. View cookie preferences.
Direct Link
Direct Link
In case you want to skip, and just grab the important details:
I start summarizing the test in the Nvidia video at about 7:20.
I start summarizing the test in the Intel video at about 7:06.
For the Nvidia tests, both systems were running the latest 331 drivers. Ubuntu had 331.13 and Windows had 331.40.
For the Intel tests, Ubuntu was using Mesa 9.3 and Windows was using the 9.18.10.3165. To my knowledge, these were the latest drivers available for both platforms.
With both these tests, we can see the difference of how a discrete graphics solution performs against an integrated solution on Ubuntu or Linux in general. They aren't representative of an entire community of hardware and configurations, but at least it should help provide us with a point of reference.
In the Nvidia tests, Dota 2 was maxed out at 1920x1080. In the Intel tests, Dota 2 was set to lowest possible settings at 1920x1080.
The tests were done on relatively some of the latest hardware available in the market, with the key components being:
CPU - Intel i7-4770 3.4GHz Haswell
iGPU - Intel HD 4600
GPU - Nvidia GTX680 GDDR5 2GB
AMD discrete and integrated graphics users may experience vastly different results. Older Intel integrated solutions may also experience very different results.
As always, please help correct me if I made any mistakes and advice is always welcome. Thank you to all who have given me corrections or advice in the past, they have helped me make better videos.
Some you may have missed, popular articles from the last month:
But can you try to compare WITH otherS desktop environmentS? Because unity seems to be the most penalizing out there, a comparison with XFCE, Openbox, LXDE or even KDE would be relevant, at least try with one please.
Do you have any demanding games in mind that can easily be benchmarked, or at least keep the timing as close as possible? I'm thinking of doing an Unvanquished test, since it seems to be a pure OpenGL title (correct me if I'm wrong). Maybe even tuxkart if I can figure out a proper testing procedure.
I would like to, but I'm trying to keep each video focused. It takes quite a bit of time to set up these tests and ensure things are as similar as possible. Took me 4 days to do this one, but I do hope I can do more tests in the future with different DEs. I did test L4D2 with Unity and LXDE. LXDE gained quite a bit of fps during heavy scenes compared to Unity, so it's true Unity is a bit of a resource hog.
Yes, it would be nice to have some results with really lightweighted WM like Openbox and without any compositor (Compton etc.).
Great tests nonetheless.
I am not sure this would actually provided any tangible difference, compiz is already set up to Unredirect Fullscreen Windows By Default, So Unity and Compiz are basically already being by passed when playing a game in full screen.
So performance should be pretty much the same either way.
I don't have ET:QW and it seems to have been removed from Steam, but I do have Quake 4. If I can figure out how to install that then I'll see how feasible it is to do a good test.
Might take a rest from doing the test videos for a couple of days though, since my hands are hurting quite a bit now.