Open vs. Closed Driver Benchmarks Of AMD's A10-7850K Kaveri On Ubuntu 16.04

[MuPuF]

It was DDR3-2133 4GB x 2. It would be mentioned in the article, but Linux still doesn't sadly allow way to expose the data nicely unless you're running as root. So only when benchmarking as root does the system table show the frequency / topology / model numbers of memory.

Yeah, I am also annoyed by this, but I guess dmidecode could be ran as suid on benchmarking machines...

[W.Irrkopf]

Linux still doesn't sadly allow way to expose the data nicely unless you're running as root. So only when benchmarking as root does the system table show the frequency / topology / model numbers of memory.

How about running a small script during system startup that collects this information and caches it somewhere in world readable format? In that case the benchmark could use this as a fallback if not running as root.

[Michael]

Look at the image with the specs: its url contains 1602217-GA-AMDKAVERI64. That's the id of the test.
Then you can go to http://openbenchmarking.org/result/1...GA-AMDKAVERI64

or if clicking on the little OpenBenchmarking.org icon on the graphs (assuming your browser is serving you the SVG version).

[Michael]

Yeah, I am also annoyed by this, but I guess dmidecode could be ran as suid on benchmarking machines...

When wiping the machines so often, etc, it's nice when everything just works out of the box.

[Michael]

How about running a small script during system startup that collects this information and caches it somewhere in world readable format? In that case the benchmark could use this as a fallback if not running as root.

It's only the system memory information via dmidecode and the Radeon clock speeds due to debugfs-only exposure (and now Nouveau too, since switching over to debugfs) where as most all other system info can be easily parsed as user. Not sure that all PTS users want an extra startup script just to record a few extra bits of data. It would be beneficial to everyone if there was a better solution upstream in the kernel.

[Dea1993]

can you test some steam games?
for example:
L4D2
CS:GO
Dirt: Showdown
Grid Autosport
Insurgency
Metro 2033/Last Light
Borderlands 2/The pre sequel
Bioshock infinite
XCOM Enemy Unknow
XCOM 2
The Witcher 2

and can you test 1080p and 720p?

thanks

[Michael]

can you test some steam games?
for example:
L4D2
CS:GO
Dirt: Showdown
Grid Autosport
Insurgency
Metro 2033/Last Light
Borderlands 2/The pre sequel
Bioshock infinite
XCOM Enemy Unknow
XCOM 2
The Witcher 2

and can you test 1080p and 720p?

thanks

The Steam-compatible games are listed here: http://openbenchmarking.org/tests/pts Most of the games you listed can't be tested since they don't properly support auto benchmarking.

However, Steam games weren't tested for this article since this is one of my daily benchmark systems in the lab where I don't have the time or space for setting up Steam there.

[reepca]

I notice the results seem a bit higher than what I typically experience in Xonotic, but then I haven't run any real benchmarks. Which reminds me, I've been meaning to ask - if my motherboard has 4 DIMM slots, and I have two sticks, and it's got two yellow and two black ones, should one stick be in black and one in yellow in order to be in dual-channel mode?

[tomtomme]

or if clicking on the little OpenBenchmarking.org icon on the graphs (assuming your browser is serving you the SVG version).

Thx!
So here the comparison of the 7850k r7 graphics kaveri with the intel hd530, for anyone interested.
http://openbenchmarking.org/result/1...GA-INTELSKYL68

I combined the two result-urls manually with an ",". Is there an easier way, a button I overlooked?

[molecule-eye]

Like most other people, I don't find the benchmarks to be very useful, since those games have been running well under the open source drivers for quite some time and since nobody plays them. Like I mean, literally, nobody. Would be nice to see some Steam games.