1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Test of CPU for Gaming : 30 CPUs compared

Discussion in 'General Hardware' started by DGLee, Apr 23, 2014.

  1. DGLee

    Joined:
    Jun 9, 2011
    Messages:
    29 (0.02/day)
    Thanks Received:
    82
    Location:
    South Korea
    I tested & measured gaming performances of "all" properly-existing CPUs in the market. Criterion for determining 'properly-existing' is that, a product shouldn't be being replaced by latter SKU at the same price level(ex: IVB is replaced by Haswell). Hope that could help those who are considering buying or upgrading one's gaming machine. Here we start.

    Test system is composed as follow:
    (+ VGA : GF GTX780Ti)
    [​IMG]

    Since the host provider (of my blog) consider sudden traffic increment as an abusing, I'm going to upload the summary part of all results only.
    Here's the performance summary at 1680 x 1050 resolution. The former one emphasizes just a sum of framerates of all games while the latter one implies average relative performance of each CPU. Of course the latter one is more important.
    [​IMG]
    [​IMG]

    Here's for 1920 x 1080 resolution.
    [​IMG]
    [​IMG]

    Here's for 2560 x 1600 resolution.
    [​IMG]
    [​IMG]
    Well, the article is over. Thanks for reading this.
    Have a nice day! :)
     
    Last edited: Apr 23, 2014
  2. night.fox

    night.fox

    Joined:
    Dec 23, 2012
    Messages:
    844 (1.33/day)
    Thanks Received:
    390
    Location:
    south korea
    thanks for sharing but it would have been nice if you throw in 3570k, 3770k, 4570k and 4770k in the comparison. This is to find out about if there is any difference in the haswell refresh CPU's
     
    15th Warlock says thanks.
  3. RCoon

    RCoon Forum Gypsy

    Joined:
    Apr 19, 2012
    Messages:
    6,879 (7.80/day)
    Thanks Received:
    3,151
    Location:
    Gypsyland, UK
    4820K is identical to the 4770K, just a different platform (which will likely have a miniscule difference to the 4790K), and there is probably miniscule difference between the 4690 and 4670.
     
    night.fox says thanks.
  4. Tsukiyomi91

    Tsukiyomi91

    Joined:
    Feb 18, 2013
    Messages:
    62 (0.11/day)
    Thanks Received:
    13
    Location:
    KL, Malaysia
    Interesting thread... and once again, Intel's somewhat affordable Core i5-4460 smoked AMD's high-end FX-9590 in FHD resolution gaming which means that it's the best bang for your buck CPU for those who're building/purchasing a pre-build system on a budget.
     
  5. Lionheart

    Lionheart

    Joined:
    Apr 30, 2008
    Messages:
    4,048 (1.74/day)
    Thanks Received:
    808
    Location:
    Milky Way Galaxy
    Smoked? o_O Massive exaggeration there and most games out there don't even take advantage of more then 2 - 4 CPU cores/threads at a time but I do agree with you that Intel's i5 series are one of the best bang for buck CPU's you can get. AMD's FX8350 compete's well price performance wise but the FX-9590 is meh IMO :D
     
  6. n-ster

    Joined:
    Jan 11, 2009
    Messages:
    8,859 (4.27/day)
    Thanks Received:
    1,405
    Tbh I don't think smoked is that much of an exaggeration. It's a pretty big difference in performance when games are supposed to be very GPU bottlenecked
     
  7. Melvis

    Melvis

    Joined:
    Mar 18, 2008
    Messages:
    3,577 (1.51/day)
    Thanks Received:
    522
    Location:
    Australia
    Clearly not smoked, more like what are you guys smoking? go look at it again. Clearly 5% isnt smoked, if you think it is then your an idiot. Also going by these results a i7 4960x isnt even double the performance of a A4-6300?? like hello!! (at 1080P) So save ya self some $$$ and buy AMD it WILL do the job.

    Oh and a A4-6300 gets the same FPS as a i7 4960 at 1080P in AVP so clearly it comes down to the games you play :slap:
     
    eidairaman1 says thanks.
  8. Vario

    Vario

    Joined:
    Oct 21, 2005
    Messages:
    2,888 (0.89/day)
    Thanks Received:
    911
    Nope 4820K is an Ivy Bridge -E (3770k) on LGA 2011. 4770k should be faster.
     
  9. eidairaman1

    eidairaman1

    Joined:
    Jul 2, 2007
    Messages:
    12,399 (4.70/day)
    Thanks Received:
    1,457
    meant for LiN/LOX modders.
     
  10. Durvelle27

    Durvelle27

    Joined:
    Jul 10, 2012
    Messages:
    2,540 (3.18/day)
    Thanks Received:
    632
    Location:
    Memphis, TN
    hmmmmm seems a little odd
     
    Crunching for Team TPU
  11. ThE_MaD_ShOt

    ThE_MaD_ShOt

    Joined:
    Nov 29, 2011
    Messages:
    4,844 (4.73/day)
    Thanks Received:
    3,387
    Location:
    Hi! I'm from the Internet
    All this proves is no matter which brand you chose, you can game just fine. Your eyes will never tell the difference.
     
    adulaamin and eidairaman1 say thanks.
    Crunching for Team TPU
  12. ELiTRiGG3R

    ELiTRiGG3R

    Joined:
    Feb 18, 2013
    Messages:
    255 (0.44/day)
    Thanks Received:
    31
    Location:
    India
    are i3 4150 and i3 3220 same in cpu power and threading?...I know that the major change from 3rd to 4th was the drastically improved intel hd igpu (don't care..better to get apu)..how about those cups would compare in gaming?I think it should be within +-3%
     
  13. Tsukiyomi91

    Tsukiyomi91

    Joined:
    Feb 18, 2013
    Messages:
    62 (0.11/day)
    Thanks Received:
    13
    Location:
    KL, Malaysia
    differences between Ivy Bridge (LGA1155) & Haswell (LGA1150) is lower TDP & improved iGPU (Iris Pro for the Haswell Core i7 chips). AMD's APU is hampered by sub-par performances despite having a discrete GPU on the same die, the FX Series are still suffering due to it's high TDP & still nowhere as fast than Intel's affordable variant of the Core i5 chips with twice the number of physical cores on a die. Lastly, I'm no fanboy of either side since I used both brands in my daily lives, but I have to say this: AMD is becoming more of a "all show, no go" company while Intel & Nvidia are fighting each other aggressively to see who's best. Using game bundles when purchasing a VGA card or CPU may boost their sales figures, but in the long run it won't help them recover when their R&D teams isn't creating a new chip on a new silicon wafer with a new architecture like Intel. If they did create it, they would have stayed in the processor & graphics card market a little more longer with some competition for Intel & Nvidia.
     
    ELiTRiGG3R says thanks.
  14. Tsukiyomi91

    Tsukiyomi91

    Joined:
    Feb 18, 2013
    Messages:
    62 (0.11/day)
    Thanks Received:
    13
    Location:
    KL, Malaysia
    also, I still remembered there's one of AMD's "comparison" with their new FirePro vs 3 year old Nvidia Quadro workstation card in a showdown is simply plain stupid. Bet some of you here knows there's a video/article about this "popular" benchmarking tests.
     
  15. Durvelle27

    Durvelle27

    Joined:
    Jul 10, 2012
    Messages:
    2,540 (3.18/day)
    Thanks Received:
    632
    Location:
    Memphis, TN
    :confused:
     
    eidairaman1 says thanks.
    Crunching for Team TPU
  16. Aquinus

    Aquinus Resident Wat-man

    Joined:
    Jan 28, 2012
    Messages:
    6,264 (6.50/day)
    Thanks Received:
    2,078
    Location:
    Concord, NH
    IPC improvement is overstated. Also the 4820k (like the 3820 I have,) has 10MB of L3 cache instead of 8MB. So between quad-channel memory and more L3 cache, I don't think you're really going to see a big difference in performance between the two. Much like how the 3820 was a little bit faster than the 2600k but a tiny bit slower than the 3770k. It will be interesting to see how graphs like these will change as newer games come out as more software leverages more cores.
     
    eidairaman1 says thanks.
  17. n-ster

    Joined:
    Jan 11, 2009
    Messages:
    8,859 (4.27/day)
    Thanks Received:
    1,405
    CPUs are supposed to have a very low impact on gaming nowadays. The fx-9590 is approximately 65% more expensive for 6% less performance. 6% on a 780 Ti is worth 40$, so price difference relative to the performance of the fx in this particular case would make the fx double the price

    Relatively though, that 6% performance is big. There's a 4.9% difference between the i5 4590 and i7 4970x, the fx is relatively worse than double that difference.

    When you put everything into perspective, to me, smoked isn't such a farfetched description after all


     
  18. eidairaman1

    eidairaman1

    Joined:
    Jul 2, 2007
    Messages:
    12,399 (4.70/day)
    Thanks Received:
    1,457
    troll much?
     
  19. TRWOV

    TRWOV

    Joined:
    Aug 11, 2011
    Messages:
    3,492 (3.08/day)
    Thanks Received:
    2,048
    Location:
    Mexico
    huge eye opener for me, those Kabinis surely are GPU limited. I was thinking about going with a 5350 for a MAME box but maybe the Sempron 3850 would be just as good.

    EDIT: disregard comment. I thought they were tested with the IGP
     
    Last edited: Apr 30, 2014
    Crunching for Team TPU
  20. Nosada

    Nosada

    Joined:
    Mar 4, 2006
    Messages:
    251 (0.08/day)
    Thanks Received:
    51
    Very nice writeup about a subject I really wish would get more attention: the necessity of high end CPU's for graphically maxed out, high resolution gaming.

    Only comment I can give is that I'd wish the low fps games would be weighed more heavily than the high fps ones. For example, if a CPU scores 30 fps in a game compared to a competitor's 20 fps, that is a big win. If the same CPU then scores a measly 120 fps in another game versus the competition's 150 fps, that would be a barely noticeable loss. However, the totals would end up suggestion that the second CPU (170 total score) would offer a better gaming experience than the first CPU (150 total score), which obviously would be a faulty assumption ...

    I'd love to get a hold of your raw data, just to see if I could implement this idea in a usable way :)
     
    n-ster says thanks.
  21. n-ster

    Joined:
    Jan 11, 2009
    Messages:
    8,859 (4.27/day)
    Thanks Received:
    1,405
    Yea that would put a lot more meaning to the results
     
  22. eidairaman1

    eidairaman1

    Joined:
    Jul 2, 2007
    Messages:
    12,399 (4.70/day)
    Thanks Received:
    1,457
    30 fps dont guarantee smooth gameplay at all


     
  23. FordGT90Concept

    FordGT90Concept "I go fast!1!11!1!"

    Joined:
    Oct 13, 2008
    Messages:
    13,554 (6.25/day)
    Thanks Received:
    3,493
    Location:
    IA, USA
    Especially with DirectX 12 in the pipeline. AAA devs are going to move to Direct3D 12 as soon as possible which pushes as much work as possible to the the GPU.


    AMD is actually doing pretty well consider the process disadvantage to the 4### Intels (32nm vs 22nm).
     
    eidairaman1 says thanks.
    Crunching for Team TPU
  24. ELiTRiGG3R

    ELiTRiGG3R

    Joined:
    Feb 18, 2013
    Messages:
    255 (0.44/day)
    Thanks Received:
    31
    Location:
    India
    ldHey man.. Been waiting for the latest 337 driver by nvidia..have read a lot of articles and they are positive and negative but really inconclusive.. I think that it must improve my bf3 mp performance greatly as it requires CPU horsepower and I lack in that area...the driver kind of improves CPU overhead..am I right? Would I receive big increase as it fixes specifically what I was worried about when I assembled my build..right now I play at medium...surprisingly aa doesn't have as big of a increase as I expected..I get 60 constant but drops to 50-45 rarely during heavy gunfights..texture at ultra 8x af
     
  25. _JP_

    _JP_

    Joined:
    Apr 16, 2010
    Messages:
    2,681 (1.66/day)
    Thanks Received:
    734
    Location:
    Portugal
    What is a module?
    What is price/performance ratio in a market segment?

    Discrete means the GPU die is in a different package, nevermind a separated die. CPU core performance is unrelated to the GPU it might have attached to it. Performance depends on your needs so it's always a case of YMMV.

    Made me reply.
     
    ELiTRiGG3R says thanks.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Share This Page