Is Windows 7 Actually Faster Than Ubuntu 10.04?

[Kano]

Usually installing a virus scanner affects boot time extremely - not that much the startup time of a game, but a bit of course too. Slow systems, especially with 1 core only and not that much ram run really much slower than when using Linux. Also lots of autostarted apps will slow down the system even more - think of the indexing for office apps... Extra system services slow down Linux boot time of course too but not that much - using desktop search engine is definitely not makeing a system faster. Those effects are not benchmarked, but you "feel" the difference. It usually does not matter if a game runs a few fps faster or not, when then it would be only interesting below 60 fps (more extreme when low fps drop below 24), but most games run much faster already with a good gpu - at least multiplattform ones.

[Shining Arcanine]

The article is only a Ubuntu vs Windows comparison, not Linux vs Windows. Ubuntu does not represent Linux. If you want to compare performance of Windows and Linux, you should choose Linux From Scratch (LFS) or Slackware.

Gentoo would be a better choice, although I think that for a proper comparison, all that would be necessary is a custom kernel. All of the tests done are graphical tests, which the default Ubuntu kernel is horribly unoptimized to do. The default Ubuntu kernel is compiled for servers. It needs to be recompiled with 1000Hz, Dynamic Ticks, Kernel Preemption and RCU Preemption to be properly optimized for a desktop environment.

[kraftman]

While Linux has long been talked about as being a faster operating system than Microsoft Windows, in 2010 is this still the case? It seems every time we deliver new benchmarks of the EXT4 file-system it's actually getting slower, recent Linux kernel releases have not been delivering any major performance enhancements for desktop users, the open-source Linux graphics drivers are still no match to the proprietary drivers, and "bloated and huge" is how Linus Torvalds described the Linux kernel last year. This is all while Windows 7 was released last year, which many view as Microsoft's best operating system release to date.

While the newest Linux kernel is much smaller then even old XP kernel it's really huge and bloated. While people say Linux is faster how often they mean graphic drivers? I bet very rarely or never. Sadly, you didn't explain why Ext 4 became "slower", you didn't say Ubuntu has debugging enabled (however, it probably won't affect games performance). Many also view Windows 7 as slow mess. Maybe some more interesting benchmarks like threading performance?

[kraftman]

Oh, it's sad there are no OS X graphic benchmarks too.

[philcostin]

When you come to benchmark Wine, remember to run it with WINEDEBUG=-all otherwise the debugging messages will seriously hamper performance.

[Apopas]

That's where windows should pwn Linux? In graphics? As far as I see Ubuntu is about equal with windows and in some tests it has the upper hand.
Very interesting is the Unigine Sanctuary demo which shows the similar capabilities of both systems.
As far as I have seen also, Fedora was over 50% faster than Ubuntu 9.10 with NVidia drivers.
All these make me to think, how could windows stand against Fedora, or even better against Gentoo? LOL

[philcostin]

We need some Haiku benchmarks

[GreatEmerald]

Nice benchmarks, although I wish there was a UT2004 test instead of OpenArena, since there already is Urban Terror which is ioquake3, and the results of OpenArena just go off the scale (I don't think the quake3 engine was built to support FPS that are even over 600 as you can see here - most engines of that time, like Unreal Engine 1, use cpu clock-related timings so high FPS doesn't really mean more than looking at glgears information), while Unreal Engine 2 is still pretty popular in games, such as BioShock 2. And UT2004 is supported on both platforms.

That aside, I'm glad that AMD ATI drivers are pretty much identical between Linux and Windows, especially on higher resolutions. Knowing how many resources are put towards Windows development of the drivers, I'm pretty surprised it went that well. That signifies that with proper native game ports (when needed), the games would play just as well on both platforms!

Also worth noting that there were gaming benchmarks carried out by Stat64 if I remember correctly, and they show that Windows 7 performs better in pretty much every game tested than Windows XP x64 (which is based on the newer Windows 2003 kernel code than the ordinary Windows XP), and seeing the statistics, Windows XP is still the most used Windows OS, so that would mean that Linux is faster in gaming than Windows XP. Of course, real benchmarks of that would be nice as well. ...in fact, why don't I just do that, after all I can just get PTS as well

[bugmenot2]

The table from page 4 is buggy.

[Veerappan]

Next was our X-Plane 9 testing courtesy of Laminar Research. As can be gathered from the results straight lines, this test is rather GPU limited with its high quality settings.

Don't mean to nitpick too much, but did you mean to say that this test was CPU limited? Flat FPS lines at varying resolutions would seem to indicate to me that the CPU was the bottleneck and not the GPU.