The free benchmark thread

Discussion in 'General Software' started by SvenBent, Jun 24, 2019.

  1. SvenBent

    SvenBent 2[H]4U

    Sep 13, 2008
    Making this thread to keep an list of free benchmarks.
    These might not bee the newest and fanciest around but they are free and decent

    7-zip -
    WinRAR -
    GeekBench - (RC5 and OGR) -
    CineBench R15 -
    CineBench R20 - Windows store
    V-ray -
    Aida64 -
    SiSoftware Sandra -

    Unigine Sanctuary (2007) -
    Unigine Tropics (2008) -
    Unigine Heaven (2009) -
    Unigine Valley (2013) -
    Unigine Superposition (2017) -

    3Dmark99 (DX6) -
    3Dmark2000 (DX7) -
    3Dmark2001 (DX8) -
    3Dmark03 (DX9) -
    3Dmark05 (DX9c) -
    3Dmark06 (DX9 SM4) -
    3Dmark Vantage (DX10) -
    3Dmark (DX11+) -

    Catzilla -

    CS:GO - Steam (Timedemo command)
    Streetfighter IV Benchmark -
    Devil May Cry 4 -
    Unreal Tournament 2004 - (Tool:
    Star Swarm Stress Test - Steam


    Crystal Diskmark -
    Atto Benchmark -
    AS SSD Benchmark -

    Bring in suggestions with a link
    Must be Free to get. and must be able to do controlled accurate repeated benchmarks aka not Fraps
    Last edited: Aug 14, 2019
    blkt and Meeho like this.
  2. SvenBent

    SvenBent 2[H]4U

    Sep 13, 2008
    - Removed Left4Dead as its not free
    - Added street fighter IV benchmark
  3. dragonstongue

    dragonstongue 2[H]4U

    Nov 18, 2008
    cheers m8 saving this for future reference

    that being said, how many of these are NOT bias towards Intel / Nvidia vs AMD (using the term accurate and repeatable referencing those 2 is like mention pigs and flying like birds in the same thought process) ^.^

    I always like the heaven benchmark myself
    Final Fantasy was/is (to best of my knowledge) only really a valid test for Nvidia's newest OR for system stability/heat testing.

    Term benchmark is "simple" that is, do quicker or less power etc then this other that set the "benchmark" unfortunately there are far too many ways for product A to do things in it's own way to make you think it is faster than product B even though it likely is not,

    I wish there were some type of "to the metal" testing, this way here we the consumer and them the hard/software vendors would be able to tell right off the bat it ABSOLUTELY is quicker for such and such reasons (because the benchmark listed out the why and what for without relying on stupid tricks to "lie" about the performance etc.
  4. SvenBent

    SvenBent 2[H]4U

    Sep 13, 2008
    There are 2 Mentalities of benchmarks
    "to the metal"

    both are valid but used for diffretn purpose. i often see people not understading the diffrent

    To the metal mentaliet

    Testing intel 9900k vs AMD 3700x in a game at 640x480 resolutions
    This test which CPU is the fastes in the given task
    Lets say CPU X is 30% faster then it is most likely 30% when all teh cpu power is neededs

    it good for technical knowledghe and peopel that can think for themsevels but its horrible as a bying advies

    Meet "effective" mentality
    Here we test them in well known used resolutions like 1440p and realide the diffrent between between the 2 cpu is now only 5%
    Its not the CPU x is suddenly slower compared to the competitor it just that in this sitatuion the cpu is just not a big of an effect.
    These kind of test are good for buying advise right here and now but they lack the proper raw numbers to tell estimate about the future

    AMD vs intel bias is not a think in "effective" benchmarks because it doesn matter the CPU x is 100% faster, if it is not their effective due to game bias or optimization path.
    To put it diffrentul. ALL and any benchmarks is important it just a question of how to proper understand and utilize the results
    Put it a third way. if CPU come out ahead wit 10% even thoug its tecnhical slower due to better code optimisation. it still the better choices for that game/engine/software. Whatever

    Ganboyisme just have to freaking die
    dragonstongue likes this.
  5. DogsofJune

    DogsofJune 2[H]4U

    Nov 7, 2008
    Last edited: Aug 6, 2019
  6. SvenBent

    SvenBent 2[H]4U

    Sep 13, 2008
    I've run into some crash issues with catzille also buts i still consider it a nice extra benchmarks

    I've added Star swarm stress test Thank you
  7. Nenu

    Nenu [H]ardened

    Apr 28, 2007
    Do you recommend a small footprint memory bandwidth/latency test util?
    Instead of installing a monolith like Sandra, just perform quick tests.
  8. SvenBent

    SvenBent 2[H]4U

    Sep 13, 2008
    I used to know a lot but I always found them very artificial. for more realworld application I tend to use 7-zip

    I'll see if i can find someone more "metal" benchmarks for ram

    Aded V-ray
    Its requires creating an account ( did mine with aa tempmail account) but its free to download
  9. SvenBent

    SvenBent 2[H]4U

    Sep 13, 2008
    added aidai64 and sisoft sandra as well