Linux Hardware, Software Trends Over One Year

[disi]

surprise surprise the radeon driver beat the FGLRX in market share!

and this number will grow because of the support drop of the hd2000-hd4000 cards.

I think that happened already a year ago in the Linux graphics survey?
The radeon driver is overall more used than the fglrx. Since Nvidia doesn't really provide an alternative and you cannot properly use xrandr with it, this is a reason for me to buy AMD

[droidhacker]

It is clear that the data has been heavily manipulated to give a false impression; though not immediately obvious, I can appreciate how the "ati" and "radeon" drivers are different, despite being the same thing for different hardware generations.... but in the second chart, where ATI and AMD are separated as graphics card vendors, it is CLEAR that the objective here was to promote nvidia rather than provide an unbiased comparison of the trends. At least the WRITTEN part treats this more correctly.... but as they say, a picture is worth a thousand words.

[Michael]

It is clear that the data has been heavily manipulated to give a false impression; though not immediately obvious, I can appreciate how the "ati" and "radeon" drivers are different, despite being the same thing for different hardware generations.... but in the second chart, where ATI and AMD are separated as graphics card vendors, it is CLEAR that the objective here was to promote nvidia rather than provide an unbiased comparison of the trends. At least the WRITTEN part treats this more correctly.... but as they say, a picture is worth a thousand words.

LOL? This is the raw data from the Phoronix Test Suite submitted to OpenBenchmarking.org.... PTS is all open-source... OpenBenchmarking.org automatically creates the graphs, there is no human involvement. PTS just reports the drivers as-is when detected.

[Vadi]

Thanks for posting the stats. The graphs for the OS vendors a least confirm that these threads have a lot of vocal minorities

I still wonder what are the system logs and such are used for when they're to be uploaded?

[Michael]

I still wonder what are the system logs and such are used for when they're to be uploaded?

- Simply for reference and checking if you didn't falsify the XML file (granted, you could have falsified the system logs, etc). Or for checking dmesg for GPU hangs, etc.

- Crowd-sourced look at some things, e.g. motherboards with broken ASPM - http://www.phoronix.com/scan.php?pag...item&px=OTk4NQ (more in terms of this crowd-sourced look are on my roadmap for the future).

- cpuinfo files, for example are analyzed to provide data like on this CPU search page listing - http://openbenchmarking.org/s/Intel%20Core%20i5-2500K (at the bottom with the example cpuinfo, etc)

- Plus a lot of other features forthcoming, but alas still battling them on my TODO list.

[Kivada]

Cirrus, Innotek and Matrox?

Innotek now only makes shock collars and in ground fence-less-fences for dogs, Cirrius hasn't made anything bought by more then 5 people in who knows how long and the last thing Matrox ever produced that anyone ever bought was the Triple Head which was made pretty useless by 6x Eyefinty cards. Sure there was the Parhelia, but it was slow and overpriced compared to the Quadro and FireGL cards it was competing with. Yeah, I know they apparently sold quite a few to run multi-monitor sets usually used in the financial sector if all the demos I've seen always having what appear to be several stock charts on the screens, which would explain a bit of the stupid in the financial sector...

[Kivada]

It is clear that the data has been heavily manipulated to give a false impression; though not immediately obvious, I can appreciate how the "ati" and "radeon" drivers are different, despite being the same thing for different hardware generations.... but in the second chart, where ATI and AMD are separated as graphics card vendors, it is CLEAR that the objective here was to promote nvidia rather than provide an unbiased comparison of the trends. At least the WRITTEN part treats this more correctly.... but as they say, a picture is worth a thousand words.

Again, this is Larabel were talking about, it's probably not even malicious, just ineptitude. Any sane person would have merged the data sets, picked their own colors for the graph and made the graphs, zoomed in on the areas in the graph sections where several data points overlap making them impossible to follow in the full sized graph and standard bars of 10 or 20 instead of 22, 38, 47 and so on...

Like I say all too often, he loves to beg for money and complain hardware companies never send him anything but he doesn't seem to be willing or able to improve the quality of his work to a point that anyone with half brain would be willing to give him anything.

And yeah I'm sure this post will be deleted by Larabel as he doesn't like direct criticism. Really wishing some other site like Ubuntu Vibes would start doing hardware reviews, they can't do much worse ad in fact are allot better at posting relevant Linux gaming news without being sensationalist or having recursive link soup.

[Milyardo]

Again, this is Larabel were talking about, it's probably not even malicious, just ineptitude. Any sane person would have merged the data sets, picked their own colors for the graph and made the graphs, zoomed in on the areas in the graph sections where several data points overlap making them impossible to follow in the full sized graph and standard bars of 10 or 20 instead of 22, 38, 47 and so on...

Like I say all too often, he loves to beg for money and complain hardware companies never send him anything but he doesn't seem to be willing or able to improve the quality of his work to a point that anyone with half brain would be willing to give him anything.

You must not be familiar with the quality of work found in the Windows equivalent of sites like Phoronix.

[Kivada]

You must not be familiar with the quality of work found in the Windows equivalent of sites like Phoronix.

I don't read Tom's if that's what you're implying lol, they don't even hide the fact their reviews are paid for and their stupid really showed in their first reviews of SSD drives where they attributed SSDs to killing battery performance after running a test that maxed out the CPU and disk non stop till the battery died, but the SSD unit did something like 3x more work units then the HDD model did before the battery ran out.

So yeah, when I read windows site I look at http://madshrimps.be --- http://overclockers.com --- http://xtremesystems.org --- http://silentpcreview.com --- http://hardocp.com --- http://hardwaresecrets.com --- http://pcper.com --- http://guru3d.com --- http://jonnyguru.com --- http://hardwarecanucks.com and a few others. Reviews in aggregate help weed out the crap.

[crazycheese]

So yeah, when I read windows site I look at .... http://overclockers.com .... http://jonnyguru.com

That two are OS-agnostic and Linux friendly.