KDE & Xfce Don't Lead To Performance Wins Over Windows 8

[Kakarott]

Thank you!

[jayrulez]

This benchmark raises some questions:
Windows 8:
- Intel Family 6 Model 58 Stepping 9
- Disk 187GB

Ubuntu:
- Intel Core i3 @ 1.8 Ghz
- 500 GB HDD + 24GB SSD

That test appeared flawed considering different specifications. Ubuntu might be one of the slow distributions and
Windows 8 was optimized for that ultrabook . I think the benchmark need a redo with the same hardware.

Did you not read the article?

[startzz]

At these resolutions, i feel confident Intel could at least get 4 or 5 fps.

Well yeah, intel hd4000 graphics should get a few fps, but its still nothing and totally unplayable, we would not be able to see performance increases and other changes with new graphics drivers, cause its not like developers will add a few lines of code, and fps in this game will increase to like 30 fps from 5

[startzz]

Hey, I thought you were only on this website so you could troll people. But you're agreeing with someone, which means you're failing as a troll.

I'm not a troll, but are a very big one

[Michael]

This benchmark raises some questions:
Windows 8:
- Intel Family 6 Model 58 Stepping 9
- Disk 187GB

Ubuntu:
- Intel Core i3 @ 1.8 Ghz
- 500 GB HDD + 24GB SSD

That test appeared flawed considering different specifications. Ubuntu might be one of the slow distributions and
Windows 8 was optimized for that ultrabook . I think the benchmark need a redo with the same hardware.

As said numerous times between the articles, it was the same exact hardware... The automated hardware/software parsing abilities on Windows is just less than PTS on Linux.

[Guest]

As said numerous times between the articles, it was the same exact hardware... The automated hardware/software parsing abilities on Windows is just less than PTS on Linux.

Too bad you didn't put different games into the benchmark as well. Two games using the same engine is a joke. The only meaningful thing from this benchmark is KDE performing slower with suspended compositions on the Intel hardware (with nvidia it should perform better). Is there any chance to see benchmarks with nvidia hardware?

[Michael]

Too bad you didn't put different games into the benchmark as well. Two games using the same engine is a joke. The only meaningful thing from this benchmark is KDE performing slower with suspended compositions on the Intel hardware (with nvidia it should perform better).

From the article: "For those not liking the selection of cross-platform games used, it's what was known to work well on both operating systems with similar renderers, e.g. using OpenGL and not DirectX and when testing with the highly-optimized proprietary AMD/NVIDIA drivers that are largely common code-bases across operating systems will yield comparable frame-rates. For the Source Engine games, there are still ongoing Linux performance optimizations and also still some Steam client problems with the automated benchmarking process for the side of it. If you feel differently about the tests, feel free to run them yourself using the Phoronix Test Suite or writing your own OpenBenchmarking.org test profiles to cover other areas of 3D performance."

Is there any chance to see benchmarks with nvidia hardware?

From the article: "If you wish to see other such cross-platform comparisons done for other hardware and more thorough analysis, please highly consider subscribing to Phoronix Premium, disabling AdBlock, and/or making a PayPal contribution. Many of the tests are very time consuming and in cases such as today's testing the hardware has to be purchased at retail when vendors aren't willing to work with us for having a Linux site dabble with Windows-loaded PCs."

[Alex Sarmiento]

So what's causing this poor performance in Linux?
Also, should'n kde should be looking for a new KWIN maintainer ?

[przemoli]

So what's causing this poor performance in Linux?
Also, should'n kde should be looking for a new KWIN maintainer ?

Years of negligence on part of Intel.

They do magnificent work NOW and in NEAR PAST, but they must to overcome years of negligence.

They are still building dev team, still building gpu drivers, etc.
On Win side Intel just put more resources.

Though I do not know if Intel on Win is capable of OpenGL ES 3.0. So here may be some unique Linux only thingy...
(Anyone from Intel OTC to check what is official stance from Intel on GLES 3.0 in Win drivers?)

Though do not ask me why they do not share code........ Even in most unfavorable scenarios there should be code that can be easily shared (like shader compilators..)

[Rexilion]

Also, should'n kde should be looking for a new KWIN maintainer ?

lol?





hereareyourcharshehe