PCLinuxOS 2010 Runs Against, But Falls Short Of Ubuntu 10.04
Phoronix: PCLinuxOS 2010 Runs Against, But Falls Short Of Ubuntu 10.04
While a bulk of the Linux community is engaged by the release of Ubuntu 10.04 LTS next week, another popular Linux distribution put out its first major update of 2010 a few days ago. PCLinuxOS, the Mandriva-based OS that is one of the top 10 most popular Linux distributions according to DistroWatch, is out with this update that bumps their kernel to Linux 188.8.131.52 while still carrying the BFS scheduler, updated NVIDIA and ATI driver support, locale improvements, and much more. We have a few benchmarks up this morning of PCLinuxOS 2010 compared to the latest snapshot of Ubuntu 10.04.
Is it just me or did the colours in the charts change?
Yeah, their colors changed last week to more match the Phoronix site's official green color. Plus in preparation for other changes on the way.
Originally Posted by MartjeB
I'm interested in where you found this "daily snapshot from 2010-04-24"
http://cdimage.ubuntu.com/daily/ only lists 19-Apr-2010?
I can't find this Ubuntu "(a daily snapshot from 2010-04-24)" http://cdimage.ubuntu.com/daily/ only has April 19. Where do you/I find this latest build?
Benchmarking desktop-oriented distributions
The results are interesting, but i think that the way Phoronix benchmarks desktop distros (in particular) can be improved.
Users probably do not mind much that their video encoding ends 1s earlier if they used distro X instead of distro Y. I still think those benchmarks can be interesting, but only if they reveal significant improvement or regression.
However, users are happy when the desktop feels more reactive overall, even if the performance in CPU-intensive benchmarks (like video encoding or raytracing) is slightly degraded. It is not an easy thing to measure, but maybe a tool like Interbech could be (revived and) included in PTS.
Any ideas/comments about the differences between PCLOS 2010 and Ubuntu 10.04 on disk performance (postmark) and LAT (graphicsmagick) ?
I don't think BFS has much to do with the disk performance difference. Perhaps PCLOS uses a different I/O scheduler or slab allocator? Any PCLOS users here?