Page 1 of 2 12 LastLast
Results 1 to 10 of 19

Thread: Testing Out The Configurable TDP On AMD's Kaveri

  1. #1
    Join Date
    Jan 2007
    Posts
    13,405

    Default Testing Out The Configurable TDP On AMD's Kaveri

    Phoronix: Testing Out The Configurable TDP On AMD's Kaveri

    AMD Kaveri APUs offer a configurable TDP to target running the APU at lower power rating and this feature can be configured on supported motherboards. While the configurable TDP is targeted for the lower-end Kaveri APUs, I ran some tests from the A10-7850K when bumping its TDP from 95 Watts to 45 to 65 Watts to see how the performance is impacted.

    http://www.phoronix.com/vr.php?view=19714

  2. #2
    Join Date
    Sep 2011
    Location
    Rio de Janeiro
    Posts
    192

    Default

    From the benchmarks, CPU performance seems to be much more constrained by the TDP than GPU. So either the GPU is extremelly efficient or it is bottlenecked by something else, which may very well be the CPU itself. Maybe overclocking the CPU will render further FPS gains in graphics benchmarks.

  3. #3
    Join Date
    Sep 2008
    Location
    Vilnius, Lithuania
    Posts
    2,395

    Default

    And what about heat and power consumption? That's kind of the point of TDP...

  4. #4
    Join Date
    Aug 2012
    Posts
    11

    Default missing Power Consumption

    I agreee with GreatEmerald, the article does seem to be missing Power Consumption...maybe an update to the article later on?

  5. #5
    Join Date
    Nov 2013
    Posts
    93

    Default

    Interesting results, Kaveri really does good when paired with a low TDP.
    It's also showed by the A8-7600, which in most benchmarks is always just 10/15% slower than the A10-7850K, while having a lower TDP and a lower clock. (Too bad that AMD didn't send you one of those too, Tje A8 should be ~$70 cheaper than the A10 when it hit the market, so a comparision under Linux would have been interesting).

    It really shows that Kaveri was optimized to work with low TDP. Or maybe it's just because it's using 28nm SHP, who knows.


    It's also interesting to see the Watt per FPS performance (results rounded, I've assumed that auto is 95W):
    Xonotic
    45W - 2.95 W/fps
    50W - 2.8 W/fps
    55W - 2.63 W/fps
    60W - 2.43 W/fps
    65W - 2.3 W/fps
    95W - 1.71 W/fps

    x264
    45W - 1.53 W/fps
    50W - 1.48 W/fps
    55W - 1.4 W/fps
    60W - 1.31 W/fps
    65W - 1.26 W/fps
    95W - 0.97 W/fps

  6. #6
    Join Date
    Nov 2013
    Posts
    93

    Default

    Quote Originally Posted by Spittie View Post
    [...]
    Damn, fps/W, not W/fps.

  7. #7
    Join Date
    Jan 2010
    Location
    Portugal
    Posts
    944

    Default

    Quote Originally Posted by Spittie View Post
    It's also interesting to see the Watt per FPS performance (results rounded, I've assumed that auto is 95W)
    Very nice! So it seems that the further you drop the TDP the more efficient it becomes. That will surely bring some very interesting laptop CPUs.

  8. #8
    Join Date
    Sep 2008
    Location
    Vilnius, Lithuania
    Posts
    2,395

    Default

    Quote Originally Posted by Spittie View Post
    It's also interesting to see the Watt per FPS performance (results rounded, I've assumed that auto is 95W):
    Xonotic
    45W - 2.95 W/fps
    50W - 2.8 W/fps
    55W - 2.63 W/fps
    60W - 2.43 W/fps
    65W - 2.3 W/fps
    95W - 1.71 W/fps

    x264
    45W - 1.53 W/fps
    50W - 1.48 W/fps
    55W - 1.4 W/fps
    60W - 1.31 W/fps
    65W - 1.26 W/fps
    95W - 0.97 W/fps
    Eh? How are you getting these numbers? Also, what is "W/fps" wouldn't efficiency be calculated in FPS/W (as in how many FPS you get per 1 watt → higher is better)?

  9. #9
    Join Date
    Dec 2012
    Posts
    362

    Default

    Yeah, I think that is fps / watt. Though it shouldn't really be that surprising that doubling tdp doesn't double performance - we have known for years we have been descending this sliding scale of performance gains by just overvolting the same silicon. That is why the original p4 failed at 200w, Nehalem capped at 150w, and Ivy Bridge could barely push 120w without hitting huge volt limitations / diminishing returns.

    Better silicon means more efficient processes, but that means you hit the roofs of performance at lower wattages.

    Good to know Kaveri is rocking it at low tdp finally - the 28nm node seems to be doing good things for AMD. Intel might finally feel a sting in their sub-50w parts.

  10. #10
    Join Date
    Jan 2009
    Location
    Outthere, NSW, Australia
    Posts
    223

    Default

    Quote Originally Posted by Figueiredo View Post
    From the benchmarks, CPU performance seems to be much more constrained by the TDP than GPU. So either the GPU is extremelly efficient or it is bottlenecked by something else, which may very well be the CPU itself. Maybe overclocking the CPU will render further FPS gains in graphics benchmarks.
    Yes, it's the RAM-bound GPU that makes the GPU look 'unaffected' from what I can tell. There was a discussion about this a couple weeks ago in another article which lead Michael to then go and do another article based on Kaveria (sorry, no link. I just use the RSS feeds) a couple days ago about this which pretty much backed up the theories (which was originally about how bound to RAM the onboard GPU's from even the early days actually were).

    It's why I've always bought the highest compatible RAM for my HTPC's and low-end machines. No point burning power that's unused because the GPU was throttled. But now we can reduce the power usgae (with proof yet to come on ACTUAL power consumption) and still get damn fine GPU-based performance. Good times all-round I say, especially in light of games not really requiring much in the way CPU performance.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •