Announcement

Collapse
No announcement yet.

How does prime test relate to gaming?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • How does prime test relate to gaming?

    Hi, i been using passmark performance tests and use the cpu test scores to compare the performance of cpu.It is very useful.

    However, my question is ,how does prime number test(one of the test run for cpu test) relate to a cpu gaming performance?
    For example, im told that floating point test is very important score for gaming.

    Overall cpu test score within variety of cpu maybe same but ,but the sub-test have different distribution.
    For example, single thread score can be very high but , multicore performance may differ and modern gaming may show a very different performance, RELATIVELY and by estimation.

    My situation:I have an Athlon 845 and few intel cpus(i5 2300,i3 4130,i3 6100 and i5 3330)
    All of this cpu have almost the same overall cpu score.
    HOWEVER, within the test, the Athlon 845 scores high on INTEGER TEST BUT LOW ON FLOATING POINT UNIT OF i5 cpu.
    Compared to i3 6100, it scores higher on integer and floating point test, and also few others, suprisingly!!!

    MAIN CONCERN
    However prime numbers test were almost twice LOWER THAN i3 6100. So, does this indicate the cpu will twice SLOWER FOR GAMING(generally) than the i3 6100?


    Extra questions, does physics test have anything to do with the gaming performance ?
    Generally AMD cpu scores lower when it comes to this test.

    On the bright side, Athlon 845 ,or any AMD generally, score higher on compression, sorting and integer performance .

    So, i hope someone could educate me.

    Im in the process of choosing between Athlon 845 and the old i5.

    Thank very much if you read all !!!!!!!!!!!!

  • #2
    You can find a description of the CPU tests here,
    http://www.cpubenchmark.net/cpu_test_info.html

    When you look at your range of CPUs don't forget about the number of cores. The i3-6100 is only dual core. But the Athlon 845 is quad core.

    The physics test has a direct relevance to gaming. The Physics Test uses the Bullet Physics Engine which is used in a lot of games.

    In games that can only use 1 or 2 CPU cores, the i3-6100 should crush the Athlon 845.
    In games that use a lot of RAM, the i3-6100 should should crush the Athlon 845 (as the i3 has DDR4 support).
    In games the use 4 cores, the CPUs will be closer in performance.



    Comment


    • #3
      Originally posted by David (PassMark) View Post
      You can find a description of the CPU tests here,
      http://www.cpubenchmark.net/cpu_test_info.html

      When you look at your range of CPUs don't forget about the number of cores. The i3-6100 is only dual core. But the Athlon 845 is quad core.

      The physics test has a direct relevance to gaming. The Physics Test uses the Bullet Physics Engine which is used in a lot of games.

      In games that can only use 1 or 2 CPU cores, the i3-6100 should crush the Athlon 845.
      In games that use a lot of RAM, the i3-6100 should should crush the Athlon 845 (as the i3 has DDR4 support).
      In games the use 4 cores, the CPUs will be closer in performance.


      Hi, ur reply was helpful and thank you, a lot.
      But from your answer i got few questions n i would really appreciate your help.Take your time.I promise it will be a worthy analysis and not a waste of time.

      1)After checking the descriptions ,i can't find sufficient details. So, if physics test has direct relevance to gaming and used in many games, may i know some examples ?
      I play games like Need for speed 2015, Deus ex, Witcher 3, Quantum Break. I don't play mini games and skyrim.

      2)To be precise,im comparing i3 6100 and athlon 845.
      In games needing 1-2 cores, i3 would be faster, though it's most probably old tittle which should still run fine on athlon(Like old call of duty n such games).

      However, in demanding games, where the performance is critical(to maintain above 30fps), quad cores were necessary.
      BY comparing physics test scores from baseline(choosen carefully by looking at various factors;prebuild system,overclocks and avoid any extreme values,etc), i3 was abt 70% faster,on average, though both cpu have same overall score.

      http://www.cpubenchmark.net/cpu.php?...+Athlon+X4+845

      https://www.cpubenchmark.net/cpu.php...00+%40+3.70GHz

      So, is it safe to assume i3 6100 will be considerably faster in games than the Athlon due to physics score?

      Just a little more pls, please read thoroughly.
      ​3.With that said, game developers do publish system requirement. Battlefield 1(the developer promised to provide relevant recommendation) recommends i5 6600k and fx 6350 as minimum.
      When i compare their physics score, it differs a lot. BUT compression score and sorting score were,surprisingly, very much the same.

      So, what is your opinion on this situation?

      Comment


      • #4
        Originally posted by David (PassMark) View Post
        You can find a description of the CPU tests here,
        http://www.cpubenchmark.net/cpu_test_info.html

        When you look at your range of CPUs don't forget about the number of cores. The i3-6100 is only dual core. But the Athlon 845 is quad core.

        The physics test has a direct relevance to gaming. The Physics Test uses the Bullet Physics Engine which is used in a lot of games.

        In games that can only use 1 or 2 CPU cores, the i3-6100 should crush the Athlon 845.
        In games that use a lot of RAM, the i3-6100 should should crush the Athlon 845 (as the i3 has DDR4 support).
        In games the use 4 cores, the CPUs will be closer in performance.


        Also, maybe im noob, but both cpu support a maximum of 2133mhz ram, athlon support ddr3 2133mhz whereas i3 supports ddr4 2133mhz.
        Does this matter a lot?

        Comment


        • #5
          DDR4 is (slightly) faster than DDR3. But the recent Intel chips have a faster memory bus. So the i3 would be faster even if both where using using DDR3. But having said that, for most games the bottle isn't the RAM (unless you are using integrated graphics)

          You can do your own research on the Bullet Physics Engine. It is widely used, but I don't have a list of games handy.


          > will be worthy analysis and not a waste of time

          For who?

          If you are only playing 3 games, then you should be looking up benchmarks for exactly those games. Nothing else matters.

          > So, is it safe to assume i3 6100 will be considerably faster in games than the Athlon due to physics score?

          For many (even most) games I would think the i3 would be slightly faster. But it isn't solely due to the physics results. Modern games are huge complex things. With many algorithms doing many different things (AI, path finding, scene generation, physics simulations, moving memory blocks around, loading files off disk, compressing textures, collision detection, sound generation, video codec, data structure manipulation of B-Trees and linked lists, sorting, internet access, etc.. etc.. etc..). And each game is different.

          Comment


          • #6
            Originally posted by David (PassMark) View Post
            DDR4 is (slightly) faster than DDR3. But the recent Intel chips have a faster memory bus. So the i3 would be faster even if both where using using DDR3. But having said that, for most games the bottle isn't the RAM (unless you are using integrated graphics)

            You can do your own research on the Bullet Physics Engine. It is widely used, but I don't have a list of games handy.


            > will be worthy analysis and not a waste of time

            For who?

            If you are only playing 3 games, then you should be looking up benchmarks for exactly those games. Nothing else matters.

            > So, is it safe to assume i3 6100 will be considerably faster in games than the Athlon due to physics score?

            For many (even most) games I would think the i3 would be slightly faster. But it isn't solely due to the physics results. Modern games are huge complex things. With many algorithms doing many different things (AI, path finding, scene generation, physics simulations, moving memory blocks around, loading files off disk, compressing textures, collision detection, sound generation, video codec, data structure manipulation of B-Trees and linked lists, sorting, internet access, etc.. etc.. etc..). And each game is different.
            Well said,thanks for explaining.
            I'm making the analysis for myself to buy best cpu for my budget.Also those games were examples.
            I found out that GTA 5 uses Bullet Physics engine. Pretty bad for the Athlon.I can see fps drop during collision and explosion from GTA 5 gameplay that uses Athlon 845 to bench.
            But this i where i have to rethink. Just like you said , i am not chosing the cpu based on physics score.
            So , i ve included some attachments.
            Games are supposed to depend on floating point performance and others as well and Athlon is considerably quicker in those cases.

            With that said, as per your suggestion, i can check benchmark of games. But the results were inconsistent or unavailable. The gameplay can be at various resolution,quality settings and gpu. Also, there are times Athlon be as quick(Asssin creed Syndicate,witcher 3,etc) as i3 6100 or slower/faster. Add in the factor that videos of Athlon were rare as it's a low budget cpu .There were also many deceiving fan-made videos, let alone videos with inaccurate(extremely slow gameplays,comparatively) results. For example some even recommends pentium against the quad core athlons.
            Afterall, most people without problem don't bother to make videos and some may have other issues with their setup(not limited to this cpu,brand,etc)

            And those not the only games i am play.

            But im getting somewhere now. Thank you for the tips again.

            Comment


            • #7
              Cpu scores
              Attached Files

              Comment


              • #8
                cpu score 2
                Attached Files

                Comment


                • #9
                  I just noticed you are running the old PerformanceTest V8 release.
                  You should use V9 instead.
                  There were significant changes. For example, V8 didn't use the Bullet Physics engine, but used the Tokamak engine instead.

                  Comment

                  Working...
                  X