(Phoronix)AMD Radeon Performance Is Incredible On Linux 3.12

[Kenmitch]

What? I thought everybody states it sucked?

http://www.phoronix.com/scan.php?page=article&item=amd_linux312_major&num=1

http://www.pcper.com/news/General-Tech/AMD-cards-get-unexpected-boost-Linux-312-kernel

Interesting results it looks like.

Guess with the upcoming steam box's it's a needed plus for AMD.

 

[OCGuy]

Isn't this just a boost over Linux 3.11?

If a driver update can increase performance 50% in a specific OS, does that mean it is performing "incredibly" now, or does that mean it was terribly gimped previously?

AMD marketing has sure been putting lipstick on some pigs recently. FYI that has the exact same meaning as the previous, for the uninformed.

Warning issued for inflammatory language.
-- stahlhart

 

[futurefields]

Isn't this just a boost over Linux 3.11?

If a driver update can increase performance 50% in a specific OS, does that mean it is performing "incredibly" now, or does that mean it was terribly gimped previously?

What is with all the turd polishing from AMD these days?

That's an incredibly stupid way to spin this topic. You can do better than that.

 

[ShintaiDK]

Its not even the driver, its the kernel. But its a pretty useless comparision without other vendors. If Intel and nVidia for example gives the same boost....

Not to mention they use a HD4xxx. AMD got severe problems with its HD7xxx driver for Linux.

 

[Kenmitch]

Isn't this just a boost over Linux 3.11?

If a driver update can increase performance 50% in a specific OS, does that mean it is performing "incredibly" now, or does that mean it was terribly gimped previously?

What is with all the turd polishing from AMD these days?

I don't really get your reasoning.

Performance increase from AMD's offerings = Turd polishing

Performance increase from NVidia = Miracle, better driver team?

Does that wrap it up nicely?

Its not even the driver, its the kernel. But its a pretty useless comparision without other vendors. If Intel and nVidia for example gives the same boost....

Not to mention they use a HD4xxx. AMD got severe problems with its HD7xxx driver for Linux.

Yep....Just the kernel made the boost. Not sure how/why but a boost is a boost.

 

[OCGuy]

I don't really get your reasoning.

Performance increase from AMD's offerings = Turd polishing

Performance increase from NVidia = Miracle, better driver team?

Does that wrap it up nicely?

No, but a performance increase that is limited to Linux, and has nothing to do with the hardware, means that it was previously under-performing, not that it is suddenly "incredible."

 

[wand3r3r]

What is with all the turd polishing from AMD these days?

 

[Gloomy]

This is just for the open-source Radeon driver, it has nothing to do with FGLRX... right?

 

[VulgarDisplay]

No, but a performance increase that is limited to Linux, and has nothing to do with the hardware, means that it was previously under-performing, not that it is suddenly "incredible."

What if it wasn't an update to their drivers that gave the increase, but a fix to the actual OS? Still AMD's problem?

 

[OCGuy]

What if it wasn't an update to their drivers that gave the increase, but a fix to the actual OS? Still AMD's problem?

That is exactly what it was it sounds like, since the article says the driver team is looking in to why it happened.

It isn't AMD's fault that it was borked, but it also doesn't mean it is incredible because it is fixed.

 

[ShintaiDK]

AMD linux driver for HD6xxx and down is actually ok. But again, why no HD7xxx numbers?

 

[Vesku]

No, but a performance increase that is limited to Linux, and has nothing to do with the hardware, means that it was previously under-performing, not that it is suddenly "incredible."

Driver performance will always be OS specific but, yes, the AMD proprietary Linux drivers are not nearly as optimized as their Windows drivers. AMD has released a lot of hardware information to the public but without some corporate backed developers Open Source work on complex software can be quite slow.

The polishing a turd comment is going overboard, imo.

 

[thilanliyan]

Isn't this just a boost over Linux 3.11?

If a driver update can increase performance 50% in a specific OS, does that mean it is performing "incredibly" now, or does that mean it was terribly gimped previously?

What is with all the turd polishing from AMD these days?

What a ray of sunshine you are in every thread you grace.

Why complain or try to find the negative aspect of better performance? In the end, better performance is better.

 

[OCGuy]

What a ray of sunshine you are in every thread you grace.

Why complain or try to find the negative aspect of better performance? In the end, better performance is better.

Nothing I said is derogatory towards a performance increase, merely the way it was presented in the OP/article. But you knew that

 

[el etro]

Need GCN cards to test results.

 

[monstercameron]

phoronix did do some research and it seems that the there was a perf boosts all around with kernel 3.12 for amd foss/closed and nvidia aswell.

here is why its better http://www.phoronix.com/scan.php?page=article&item=linux_312_performance&num=1

foss vs catalyst http://www.phoronix.com/scan.php?page=article&item=linux_312_catalyst&num=1

just for fun http://www.phoronix.com/scan.php?page=news_item&px=MTQ4NzM

this kernel is great actually.

 

[DarkKnightDude]

Maybe they're targeting gamers now, thanks to valve?

 

[dacostafilipe]

Nothing to do with AMD ... it just a change in the Kernel that let the CPU stay more time in a hight frequency state.

A similar thing could also explain the BF4 Beta performance difference between Win7 and Win8 ...