First Look: AMD Trinity APU, Linux Already Runs Well

[AnonymousCoward]

No UVD, no good dynamic power management, suboptimal 3D performance. Mature support is really not the right term to use here. I think basic support is the more appropriate term. If you are using the term mature support you are setting expectations that the drivers can't fulfill, this will needlessly leave people disillusioned.

[duby229]

Alright, so how about completely stable and bug free support then? With new features being added at an amazing rate considering the number of developers working on it.

[pingufunkybeat]

I think that basic support is a good way to put it, but I'd also call it mature basic support, considering how long it's been stable and problem-free. I mean, the drivers can run Tropics and Heaven, ffs, they are doing fine. (with some OpenGL4 functionality missing, of course)

UVD and dynamic powersaving would be nice to have, I agree.

Anyway, this chip looks very interesting. Many cores, great integrated graphics. I'm not looking for a new machine at the moment, though, my current Phenom II is still kicking ass.

[Raven3x7]

Well, I'm not really expecting much. After all each int pipe is only 2 issues wide, and both int pipes in the module have to share the same frontend, which will be a bottleneck. Not to mention that most benches are going to be run on windows 7 which has a known bug that doesnt work well with BD. In addition for some god forsaken reason the big sites seem to prefer single threaded cookie cutter benchmarks.

Well actually that's not accurate. It can do 2 int ops and 2 mem ops while Phenom can do 3 ops but that's mem and int combined. I'd wait for some final silicon benches before making any conclusions although current leaks paint a rather dire picture.

[duby229]

I see. Still doesnt make much sense tho, alot less then half of the instructions will be mem ops though. Additionally Phenoms issue ports handled fp instructions too, but I am under the impression that BD new flexfp unit has its own dedicated front end which has an unknown number of issue ports.

Have you seen any good block diagrams of the frontend yet?

[Raven3x7]

I see. Still doesnt make much sense tho, alot less then half of the instructions will be mem ops though. Additionally Phenoms issue ports handled fp instructions too, but I am under the impression that BD new flexfp unit has its own dedicated front end which has an unknown number of issue ports.

Have you seen any good block diagrams of the frontend yet?

Still the most detailed i've seen http://realworldtech.com/page.cfm?Ar...2610181333&p=7

[Fenrin]

... I think basic support is the more appropriate term. ...

I would say Vesa driver basic support, Radeon driver advanced support, AMD's proprietary driver almost mature support
I'm currently happy user of a ATI card with Radeon drivers

[skotadopsyxos]

Isn't openbenchmarking submission supposed to be anonymous????

This weekend while browsing OpenBenchmarking.org when deciding what features to work on next for this collaborative testing platform that's integrated into the Phoronix Test Suite, I stumbled upon more data from some engineering samples that were submitted in recent weeks. It happened to be just like the discovery of the dual-Interlagos results earlier this year -- when an AMD government partner was running the Phoronix Test Suite and knowingly decided to share their Linux benchmark results with the world using OpenBenchmarking.org. With Bulldozer now coming to market, AMD's partners are now already looking towards Trinity. From what I have found so far, the Trinity data at hand comes from two AMD partners this time and also one system that's even traced back to Sunnyvale, California.

an ip address means it's not anonymous at all!
Am I the only one concerned about this???

[89c51]

an ip address means it's not anonymous at all!
Am I the only one concerned about this???

yes

[deanjo]

an ip address means it's not anonymous at all!
Am I the only one concerned about this???

Well it is anonymous as far as not requiring a sign in. If they were really worried about their identity they shouldn't have submitted to an online database in the first place. That is just common sense. When it comes to the internet there is no such thing as true anonymity.