EVGA

Noob question: Why do benchmarks...

Author
samnew2288
New Member
  • Total Posts : 84
  • Reward points : 0
  • Joined: 2016/08/31 15:43:23
  • Status: offline
  • Ribbons : 0
2016/09/28 17:44:30 (permalink)
Why do benchmarks look worse than actual games but run also way worse, too?
Like, I think Battlefield or other graphic intensive games looks better than most (not all) benchmarks
but benchmarks run at such low FPS?


Why is this?
 
 
Edit:  Also, I noticed that GTA V in benchmark mode runs at half the FPS speed as the actual game itself.  But, doesn't look any better than the game in my opinion, and sometimes the game looks even better.
 
 
Another edit:  Castlevania Lords of Shadow 2 looks way better than the FireStrike demo (in my opinion) and runs at a constant 50fps on my 750 Ti, but the firestrike demo/benchmark has difficulty staying about 40-50 with my 1080.
post edited by samnew2288 - 2016/09/28 17:47:43
#1

14 Replies Related Threads

    AzN-SoLjA
    CLASSIFIED ULTRA Member
    • Total Posts : 7982
    • Reward points : 0
    • Joined: 2011/06/18 11:45:02
    • Location: Kent, WA - Use My Affiliate Code: 3MHCIBME48
    • Status: offline
    • Ribbons : 14
    Re: Noob question: Why do benchmarks... 2016/09/28 18:54:12 (permalink)
    Not my expertise however, benching is to stress test rendering 3D objects, complexity of detail, shadow, anti-aliasing and etc.
    Just cause something else looks better doesn't mean its actually fully utilizing the GPU capabilities. More or less we use the benchmarks to test our overclock for stability.

    | EVGA Z68 FTW | EVGA GTX 780 Ti SC 3GB | EVGA GTX 1080 Ti FTW3 | Custom EK Water Loop | Red Harbinger Cross Desk
    | i7 2600K @ 4.5GHz @ 1.24v | Corsair 16GB Vengeance | Corsair Neutron 256GB SSD | Corsair AX850 | Lamptron FC5V3 | ASUS VE278 27" |
    | Creative Inspire T6160 5.1 & X-Fi Fatal1ty 7.1 Champ | Razer Marauder & Spectre & Tiamat 7.1 | Affiliate Code: 3MHCIBME48 | Associate Code: 08HWLPJXE0O5MPM
     
    #2
    samnew2288
    New Member
    • Total Posts : 84
    • Reward points : 0
    • Joined: 2016/08/31 15:43:23
    • Status: offline
    • Ribbons : 0
    Re: Noob question: Why do benchmarks... 2016/09/28 19:32:58 (permalink)
    That is...

    sort of what I thought.
    Ok, thank you very much for clarifying and adding.

    So, for example.  In GTA 5 the game (not benchmark)... the game might diminish textures at a long distance since they aren't being used up close and serve no purpose for the gamer.
    But in a benchmark, those textures could be left at full quality even at a long distance.   Something like that?
    I'm not sure on the terminology of that:  What's it called when a texture degrades automatically at a distance to conserve performance?

    So, in other words... there's a lot going on in a benchmark, more than meets the eye sort of thing?
    #3
    Nereus
    Captain Goodvibes
    • Total Posts : 18924
    • Reward points : 0
    • Joined: 2009/04/09 20:05:53
    • Location: Brooklyn, NYC.
    • Status: online
    • Ribbons : 58
    Re: Noob question: Why do benchmarks... 2016/09/28 19:36:21 (permalink)
     
    Also giant epeen.
     
     


      BUILD 1 2   |   MINI-ITX BUILD   |   MODSRIGS $1K WIN   |   HEATWARE 111-0-0   |   ASSOCIATE CODE CSKKXUT5Q9GVAFR

    #4
    stalinx20
    CLASSIFIED Member
    • Total Posts : 4977
    • Reward points : 0
    • Joined: 2009/01/03 08:56:23
    • Location: U.S., Michigan
    • Status: offline
    • Ribbons : 0
    Re: Noob question: Why do benchmarks... 2016/09/28 21:39:59 (permalink)


    EVGA X79 Dark
    2080 Black edition
    980
    EVGA 1000 gold PSU (Gold)
    4820K CPU
    16x G-skill
    #5
    samnew2288
    New Member
    • Total Posts : 84
    • Reward points : 0
    • Joined: 2016/08/31 15:43:23
    • Status: offline
    • Ribbons : 0
    Re: Noob question: Why do benchmarks... 2016/09/29 12:11:50 (permalink)
    Lol Nereus, what does that even have to do with anything?

    Are you saying that the programmers do this on purpose so people can feel like they have um.. giant.. um... yea..

    I think I'm guilty of this, because when I heard that NVIDIA is going to be announcing a new 12GB (titan similar) version of the 1080.  I tried to tear the article up looking for ways in that mine was better lol.

    And, for those of you who would like to know:  It sounds like NVIDIA is trying to make a Titan X version of the 1080.
    And, they're labelling it as "The power of Titan X for a much lower price"

    But, wait wait wait wait just a minute...
    The FE and other versions of the 1080 were already more powerful than the Titan X in benchmarks.
    And the FTW completely rocked the FE, other versions AND Titan X.

    So, don't they mean to say they're releasing something *better* than the Titan X?
    I confused...
    #6
    bg8780
    CLASSIFIED Member
    • Total Posts : 2540
    • Reward points : 0
    • Joined: 2008/02/19 14:21:34
    • Status: offline
    • Ribbons : 4
    Re: Noob question: Why do benchmarks... 2016/09/29 12:21:16 (permalink)
    I use them to make sure my hardware is configured properly. After installing my new 1080 I wanted to make sure it wasn't failing or having stability issues. It's also a good way to check for bottlenecks. If my scores don't stack up with what everyone else is getting, I'll know something is wrong.
    #7
    samnew2288
    New Member
    • Total Posts : 84
    • Reward points : 0
    • Joined: 2016/08/31 15:43:23
    • Status: offline
    • Ribbons : 0
    Re: Noob question: Why do benchmarks... 2016/09/29 12:33:45 (permalink)
    One of the users went into a little bit of detail on why benchmarks run slower but are less appealing to the eye than games.

    Do you have any more on that, BG8780?
    I think I'm getting it.  In a benchmark, they just crank everything way up... even stuff that is crunching numbers that you can't see.
    #8
    Sean1976
    Superclocked Member
    • Total Posts : 222
    • Reward points : 0
    • Joined: 2016/07/03 07:44:58
    • Location: Does it matter?
    • Status: offline
    • Ribbons : 0
    Re: Noob question: Why do benchmarks... 2016/09/29 12:39:13 (permalink)
    Some benchmarks are worthy and its usually the free ones that are good and worth the time. 3Dmark on the other hand sucks nuts, I paid for it some time ago and now they want me to pay again for stupid TIMESPY...... BLOW ME Futuremark!
    Your right, games like Mafia II, BF3, BF4, Crysis 3, Metro Last Light all are better examples of burn tests on cpu/gpu. Metro LL has one of the best benchamrks in its file system, if you browse local files.
     
     
    post edited by Sean1976 - 2016/09/29 12:42:19

    Intel i5 8600k @5.3ghz with 1.450 Vcore
    Asus ROG Strix-E GamingZ370 deck
    Hyper 212 Evo Air After-Cooler
    Evga 1080 GTX Founder's Edition@ 1987mhz core/5400mhz memory
    G-Skill TridentX 3600mhz XMP4000mhz (2x8GB RGB)
    1x Intel 540 series SSD 480GB
    2x Intel 520 series SSD 240GB each(480GB)
    2x Intel 535 series SSD 240GB each(480GB)
    Corsair HX850 Watt Gold Psu
    Antec Eleven Hundred v2 Carbon Fiber case w/ 9x 120mm fans
    BenQ XL2420TE 144hz 1ms Lightboost/MotionBlur/ Gaming panel(1080p)TN
    G15/G9 Logitech gear
     [URL=http://s1273.photobucket.com/user/SeanJ76/media/8600k%20new%20pics_zps2k125brl.jpg.html][/URL]
     
     
     
    #9
    samnew2288
    New Member
    • Total Posts : 84
    • Reward points : 0
    • Joined: 2016/08/31 15:43:23
    • Status: offline
    • Ribbons : 0
    Re: Noob question: Why do benchmarks... 2016/09/29 13:37:00 (permalink)
    Wow, thanks Sean.  I will check it out.
    Mafia II used to be my benchmarker actually, but.. it's a little dated now and I use GTAV as a "virtual city" benchmarker.
    #10
    Cool GTX
    EVGA Forum Moderator
    • Total Posts : 30996
    • Reward points : 0
    • Joined: 2010/12/12 14:22:25
    • Location: Folding for the Greater Good
    • Status: offline
    • Ribbons : 122
    Re: Noob question: Why do benchmarks... 2016/09/29 13:57:35 (permalink)
    Welcome to the Forums, samnew2288
     
    1) Benchmark software is a standard to which any PC can run; so that they have an overall point of comparison, even thought the PCs can be quite different or even identical in both hardware and software.
     
    2) You change the OS, drivers, hardware, OC settings ---> did you improve the performance of your PC or hamper it ? Run a benchmark.
     
    3) Is your PC setup stable ?  Benchmarks are usually synthetic and do Not behave like standard code in actual activities like gaming or Folding.
    Your benchmark best run; may in fact not even work in any specific game, forcing you to detune your OCs.
     
    (edit typo)
     
     
    post edited by Cool GTX - 2016/09/29 20:32:12

    Learn your way around the EVGA Forums, Rules & limits on new accounts Ultimate Self-Starter Thread For New Members

    I am a Volunteer Moderator - not an EVGA employee

    https://foldingathome.org -->become a citizen scientist and contribute your compute power to help fight global health threats

    RTX Project EVGA X99 FTWK Nibbler EVGA X99 Classified EVGA 3080Ti FTW3 Ultra


    #11
    sethleigh
    SSC Member
    • Total Posts : 796
    • Reward points : 0
    • Joined: 2015/08/12 11:27:56
    • Status: offline
    • Ribbons : 4
    Re: Noob question: Why do benchmarks... 2016/09/29 14:37:32 (permalink)
    samnew2288
    Lol Nereus, what does that even have to do with anything?

    Are you saying that the programmers do this on purpose so people can feel like they have um.. giant.. um... yea..

    I think I'm guilty of this, because when I heard that NVIDIA is going to be announcing a new 12GB (titan similar) version of the 1080.  I tried to tear the article up looking for ways in that mine was better lol.

    And, for those of you who would like to know:  It sounds like NVIDIA is trying to make a Titan X version of the 1080.
    And, they're labelling it as "The power of Titan X for a much lower price"

    But, wait wait wait wait just a minute...
    The FE and other versions of the 1080 were already more powerful than the Titan X in benchmarks.
    And the FTW completely rocked the FE, other versions AND Titan X.

    So, don't they mean to say they're releasing something *better* than the Titan X?
    I confused...

    You are indeed confused. NVidia hired Donald Trump's PR firm, seeing how marvelous a job they are doing making their client look good and make wise decisions. They advised NVidia to name their new flagship product the exact same thing as their former flagship product.
     
    So users have taken it into their own hands to disambiguate, by referring to the new Titan X that is based on the Pascal architecture as the Titan XP, and the old Titan X based on Maxwell simply Titan X, or sometimes I've seen Titan XM or whatever.
     
    But no, the 1080 FTW doesn't rock the new Titan XP.
     
    As to the OP, the benchmarks are designed to test the ability of the gpu using all of its features and really complex environments and details, textures, geometries, what have you. You can't predict game performance from these benchmark results, other than to say if you can raise your benchmark results, you'll probably also see a rise in game performance. That could depend on a lot of things. For example, if you raise your benchmark score by raising your cpu performance, that may not translate into higher fps in a game that isn't cpu bound. But if you raise your benchmark score by overclocking your gpu, you'll probably see a rise in game fps for games that are gpu bound, which is most of them.

    Happy EVGA customer.  Affiliate Code: 0Y7-1VU-ATW2
     
    GigaByte X570 Aorus Master, AMD Ryzen 5900x under Optimus Foundation block, 32gb G.Skill DDR4 @ 3800 MHz 14-14-14-28, EVGA 3080ti FTW3 Ultra under Optimus block, 2TB 980 Pro SSD, EVGA Supernova G6 850W PS, ASUS 34" 3440x1440p 120Hz ultrawide, Lenovo 24" 1080p secondary monitor, Win 10

    #12
    Sajin
    EVGA Forum Moderator
    • Total Posts : 49167
    • Reward points : 0
    • Joined: 2010/06/07 21:11:51
    • Location: Texas, USA.
    • Status: offline
    • Ribbons : 199
    Re: Noob question: Why do benchmarks... 2016/09/29 17:54:17 (permalink)
    samnew2288
    Why do benchmarks look worse than actual games but run also way worse, too?
    Like, I think Battlefield or other graphic intensive games looks better than most (not all) benchmarks
    but benchmarks run at such low FPS?


    Why is this?
     
     
    Edit:  Also, I noticed that GTA V in benchmark mode runs at half the FPS speed as the actual game itself.  But, doesn't look any better than the game in my opinion, and sometimes the game looks even better.
     
     
    Another edit:  Castlevania Lords of Shadow 2 looks way better than the FireStrike demo (in my opinion) and runs at a constant 50fps on my 750 Ti, but the firestrike demo/benchmark has difficulty staying about 40-50 with my 1080.


    Benchmarks look good to me depending on what settings/resolution they are run at. Official benchmarks like firestrike run at predefined settings that cannot be changed unless you do a custom run which will allow you to change all types of different settings to make the benchmark look better.
    #13
    Nereus
    Captain Goodvibes
    • Total Posts : 18924
    • Reward points : 0
    • Joined: 2009/04/09 20:05:53
    • Location: Brooklyn, NYC.
    • Status: online
    • Ribbons : 58
    Re: Noob question: Why do benchmarks... 2016/09/29 22:11:41 (permalink)
    Nereus
    Also giant epeen.
    samnew2288
    Lol Nereus, what does that even have to do with anything?

    Are you saying that the programmers do this on purpose so people can feel like they have um.. giant.. um... yea..


    LOL nope, well ..yes I guess. I meant people like to get their scores as high as they can by overclocking, and then compare scores with others... thus, the higher the score, the more giant the epeen. 
     


      BUILD 1 2   |   MINI-ITX BUILD   |   MODSRIGS $1K WIN   |   HEATWARE 111-0-0   |   ASSOCIATE CODE CSKKXUT5Q9GVAFR

    #14
    stalinx20
    CLASSIFIED Member
    • Total Posts : 4977
    • Reward points : 0
    • Joined: 2009/01/03 08:56:23
    • Location: U.S., Michigan
    • Status: offline
    • Ribbons : 0
    Re: Noob question: Why do benchmarks... 2016/09/29 22:58:26 (permalink)



    EVGA X79 Dark
    2080 Black edition
    980
    EVGA 1000 gold PSU (Gold)
    4820K CPU
    16x G-skill
    #15
    Jump to:
  • Back to Mobile