Windows 8 Outperforming Ubuntu Linux With Intel OpenGL Graphics

[Pawlerson]

Alright, so you decided to do these benchmarks using the obviously most relevant and interesting titles for PC gaming right now available, such as Nexuiz, Xonotic, and OpenArena, and by no means uninteresting and irrelevant titles such as Team Fortress 2, Unigine OilRush, Serious Sam 3, Counter Strike: Source and the like.

And the most uninteresting and irrelevant graphic chip.

[Pawlerson]

All the newer games you listed run on DirectX under Windows. How should he have compared the openGL performance between the OSes this way??

It doesn't matter if they're using OpenGL or directx. Only performance matters in this case. However, to see really fair benchmarks we probably have to wait for TF2 benchmarks running on nvidia hardware. Btw. I wonder how much $$ Phoronix gets from sponsored titles by craple and m$?

[FourDMusic]

The increase in frame rate for Nexuiz looks like MSAA was turned off rather than a performance improvement - I don't know for sure. The change in performance does match earlier benchmarks for when MSAA was not implemented in mesa to when it was in an earlier article: http://www.phoronix.com/scan.php?pag...sa91_ivb&num=2.

It would be good to have some sort of flag <in> the benchmarks showing whether MSAA is enabled/used so that we're not confusing performance regressions with MSAA being enabled and performance improvements with it being disabled. Or having it always disabled all together.

[birdie]

It's quite possible Linux/Open Source Intel drivers will never reach the performance level of Windows drivers, because the latter are close source, thus Intel can secretly cheat in Windows but they cannot cheat in Linux, 'cause their secrets/hacks/subterfuges will be instantly unearthed by their competitors and journalists.

[BO$$]

It's quite possible Linux/Open Source Intel drivers will never reach the performance level of Windows drivers, because the latter are close source, thus Intel can secretly cheat in Windows but they cannot cheat in Linux, 'cause their secrets/hacks/subterfuges will be instantly unearthed by their competitors and journalists.

Exactly. They cannot use stuff that would reveal inner workings of hardware. So they have to use the 'honest' interfaces with lower performances. Another argument that open source drivers will never be fast enough to compete with windows and closed source.

[schmidtbag]

It's quite possible Linux/Open Source Intel drivers will never reach the performance level of Windows drivers, because the latter are close source, thus Intel can secretly cheat in Windows but they cannot cheat in Linux, 'cause their secrets/hacks/subterfuges will be instantly unearthed by their competitors and journalists.

I've thought this myself.

I do think another test to replace either Xonotic or Nexuiz would've been nice, since in the end we're basically looking at only 2 benchmarks. But, I'm sure Michael will post more tests.


I would say that Ubuntu with Unity used to be the best baseline for measuring linux performance since it was the most popular and therefore the standard (whether we like it or not), but it isn't the most popular distro anymore so I'd say Mint should be the main distro to comparisons with. I guess the only problem with Mint is I'm not aware of it supporting anything other than Cinnamon and XFCE, though I haven't taken a look at it in maybe 3 years so it's hard to tell.

As for people whining about tests not used with your personal preferred game - keep in mind Michael is doing tests that his benchmark suite comes with. I highly doubt he's licensed to add whatever the hell he wants to it.

[JS987]

And the most uninteresting and irrelevant graphic chip.

Intel HD 4000 is fastest GPU with non-crappy open source drivers.

[Ancurio]

It doesn't matter if they're using OpenGL or directx. Only performance matters in this case. However, to see really fair benchmarks we probably have to wait for TF2 benchmarks running on nvidia hardware. Btw. I wonder how much $$ Phoronix gets from sponsored titles by craple and m$?

Wrong. This was a driver comparison, that's why it's crucial that the games run the same openGL codepaths under both OSes.
Otherwise factors such as better DirectX implementation of the same game would have tainted the benchmark.

[JS987]

Alright, so you decided to do these benchmarks using the obviously most relevant and interesting titles for PC gaming right now available, such as Nexuiz, Xonotic, and OpenArena, and by no means uninteresting and irrelevant titles such as Team Fortress 2, Unigine OilRush, Serious Sam 3, Counter Strike: Source and the like.

It doesn't make sense to benchmark using games which are designed for discrete GPU on Intel GPU. If you are serious gamer you will need discrete GPU anyway.

[przemoli]

People start thinking for a moment..

1) Valve DO NOT ALLOW for benchmarks of Steam nor games on Steam for Linux... RTFE(ULA). Its beta. So no nice TF2 benchmarks in near future...

2) Valve support UBUNTU, in STOCK configuration. Benchmarking stock Ubu vs stock Win8 make sense.

3) Installing another DE/tweaking Unity/Compiz is easy, so people wanting to get most perf will do it. But general benchmarks should be useful for general audience :P

4) You can rerun those tests on your configurations, and upload to openbenchmarking.org Show the results, instead of molesting poor keyboards