Radeon HD 7970 CPU scaling performance review [Guru3D.com]

Discussion in 'Frontpage news' started by Guru3D News, Dec 23, 2011.

  1. alanm

    alanm Ancient Guru

    Messages:
    12,270
    Likes Received:
    4,472
    GPU:
    RTX 4080
    There definitely IS the phenomenon of CPU bottlenecking. Its just not as bad as it was a few years ago when most people were on lower res screens. However, it can still be seen in certain games even at 1920x1200 as this review has shown (Anno 1404, and to a lesser extent FC2). For most it will not be felt, esp if over 60 fps, just as you will not feel any difference between a 7970 and a 6870 in many games.

    I have a feeling several other (not tested) games like Skyrim, Starcraft 2, GTA4 may show similar results to Anno 1404.

    Its probably unanimous that multi-GPU bottlenecking occurs more severely depending on the CPU.
     
  2. Redemption80

    Redemption80 Guest

    Messages:
    18,491
    Likes Received:
    267
    GPU:
    GALAX 970/ASUS 970
    Well bottlenecking is the trendy new word people spit out when someone has a new GPU and not the top CPU.

    Not saying it doesn't exist, it does and certain games will clearly show it does, but for present games something like an 2500k is as good as it gets for games considering the low price of it.

    Maybe next year extreme CPU's will be utilised more, proper use of DX11, more advanced AI, or even PhysX 3.x, but at present spending any more than $250 on a CPU is a waste if all your are doing is gaming.
     
  3. Norvekh

    Norvekh Guest

    Messages:
    2,677
    Likes Received:
    12
    GPU:
    EVGA RTX 2080 Ti
    I guess I'm in an interesting position. I've run the gamut, E6600, Q6600, PII 965, and now my 2600K. Each time, even at stock speeds, I've seen games run smoother even when retaining the same graphics card between processor upgrades. While the PII 965 did get similar average framerates to my 2600K, it was far more prone to dropping while playing. Now, this is partly because of the games I enjoy most which are Source games, notorious for being CPU limited. Others like Fallout 3 and New Vegas just don't stutter like they did. And GTA4 is a new game really. My minimum framerates went from the mid-teens to the upper 20s making the whole experience so completely different and so much more enjoyable. I noticed similar gains when I was running 4870 Crossfire and went from my Q6600 to my PII 965. It indeed was a fantastic upgrade. But, would I want to game on a Phenom II or AMD FX system today after having tasted Sandybridge? Absolutely not. The games are so much more consistent now and that's something that's just impossible to show without a graph of the framerate over time to show the quick dips and peaks.
     
  4. ---TK---

    ---TK--- Guest

    Messages:
    22,104
    Likes Received:
    3
    GPU:
    2x 980Ti Gaming 1430/7296
    That's where graphs with minimum framerates come into play. That's much more important than average to me.
     

  5. Redemption80

    Redemption80 Guest

    Messages:
    18,491
    Likes Received:
    267
    GPU:
    GALAX 970/ASUS 970
    That is very true, the 2600k is one of the best upgrades i have ever gotten.
    While many games didn't improve much, if at all, others felt like an entirely new engine.

    When i went form 260-460 i got a small but good upgrade, when i went from 460-570 i barely noticed it, but when i upgraded the QX9650 to a 2600k the 570 jumped into life.

    GTAIV became playable for the first time, and Capcom games just about doubled in performance, which might seem overkill at 100-200fps but if your into 3D, multi monitor gaming, or even just have a single 120hz screen, it makes all the difference.
     
  6. AbjectBlitz

    AbjectBlitz Ancient Guru

    Messages:
    3,463
    Likes Received:
    2
    GPU:
    R390 1200/1720
    Would like to see the same test done with crossfire to stop the GPU bound situations ;)
     
  7. Neo Cyrus

    Neo Cyrus Ancient Guru

    Messages:
    10,793
    Likes Received:
    1,396
    GPU:
    黃仁勳 stole my 4090
    I would.
     
  8. IcE

    IcE Don Snow

    Messages:
    10,693
    Likes Received:
    79
    GPU:
    3070Ti FE
    Why do you insist on seeing proof of something you already know? Of course slower CPU's are going to have lower minimums, it's common sense.

    Maybe this will finally stop the nonsense "LOL BOTTLENECK" we see around in every thread.
     
  9. ---TK---

    ---TK--- Guest

    Messages:
    22,104
    Likes Received:
    3
    GPU:
    2x 980Ti Gaming 1430/7296
    why did you insist on flaming me when all I wanted to see was a review with minimum framerates? merry xmas bud:)
     
  10. Matt26LFC

    Matt26LFC Ancient Guru

    Messages:
    3,123
    Likes Received:
    67
    GPU:
    RTX 2080Ti
    Think the Anand review had some min framerates
     

  11. IcE

    IcE Don Snow

    Messages:
    10,693
    Likes Received:
    79
    GPU:
    3070Ti FE
    Lol, because I was truly "flaming". Funny.
     
  12. benqbiggis

    benqbiggis Member

    Messages:
    42
    Likes Received:
    0
    GPU:
    SLI GTX 580
    why CPU is tested in BF3 single player ? very useless

    test in multiplayer and you gonna see a huge difference
     
  13. J. Doe

    J. Doe Guest

    Absolutely. I'd like it if you can especially give attention to i7 870 and BD OC'ing. I know you have a Classy 200 (I also had a 870 on one), and am wondering how it'd fare against the FX-8150. I've told people that the 870 would beat BD since it performs very similar to Nehalem per-clock, and it did the review.

    http://www.anandtech.com/bench/Product/107?vs=46

    What I'm wondering is, how it'd do against BD once both are OC'ed. The 870 is limited by multi, but taking it to 200 bCLK shouldn't be an issue with the Classy. OTOH, BD would have a higher OC'ing potential but it has worse IPC.

    It'd also be helpful if you add in some heat/power consumption values. On a H100? Because if you consider those, then the Lynnfield would come on top of BD. Only if Intel hadn't killed that platform... it still would have been a solid alternative I think.
     

Share This Page