Rydermark Benchmark Preview

Others/Miscelleneous by jmke @ 2007-05-16

Rydermark is a new 3D benchmark software created by Candella Software. We take a look at what it looks like, what options you can set, what it tests and provide you with a run through of the benchmark with a small movie.

  • next



Candella Software has been developing a 3D benchmark for the last year (or more) and was caught in the public spotlight when The Inquirer posted info about rendering precision being less on NVIDIA hardware. Since than screenshots have been posted, which in turn were claimed as fake by others.

Madshrimps (c)

Today we got a chance to run the Rydermark benchmark ourselves on a modest PC config; there are plenty of options to be set.

Madshrimps (c)

The minimum requirements for the Rydermark benchmark are set at:

  • 3.0GHz Single Core CPU
  • 1Gb system RAM
  • DirectX 9-class GPU with 128Mb RAM
  • Sound card with support for DirectSound 3D
  • Windows XP & DirectX 9.0c

    For the time being Rydermark is a Windows XP/DX9 only benchmark, maybe DX10 in the future under Vista.

    Rydermark comes with an interesting readme which gives a bit of background information on the benchmark:

  • Can we relate the benchmark performance to actual games performance?

    Absolutely. RyderMark is a technology demonstration of a racing game currently in development. It is NOT a synthetic benchmark. Just like the game, in addition to the GPU, the benchmark also stresses the CPU, audio and memory subsystems.

    The graphics is powered by Sahara, our in-house DirectX 9 graphics engine. Audio is provided by FMOD Sound System (www.fmod.org) which has already been used in dozens of games titles. Therefore, every technology used in the benchmark is directly relevant to the game.

  • Do you have different shaders for GPUs from different manufacturers?

    ABSOLUTELY NOT. This is something we wish to make crystal clear. We run IDENTICAL shader code on GPUs from all hardware manufacturers.

    To reiterate the point, one of the main reasons we can stake the claim for creating a truly impartial and fair benchmark is the fact that we do not favor any hardware manufacturer over another by supporting obscure or optional texture formats or features in DirectX.

  • How is the final RyderMark score calculated?

    The final benchmark score is calculated as follows:

    RyderMark Score = 20X Maximum Frame Rate + 40X Minimum Frame

  • Why do you value the minimum frame rate much more than the maximum frame rate?

    From our perspective as a games developer, we place a very high value on smooth gameplay. This requires a minimum of 30 frames per second at all times. At the same time, peak frame rates over 60 frames per second make no discernible difference to the gameplay experience.

    Therefore, we urge users to look closely at the minimum frame rate as well as the overall score. PCs which fail to deliver a consistent minimum 30 frames per second may not provide an enjoyable gaming experience in next generation games.

  • Can you describe some of the different shaders and technologies used in RyderMark?

    These are some, but not all of the techniques used in the benchmark:
    - Parallax Occlusion Mapping
    - Motion blur
    - Depth of field
    - Soft shadows
    - 32-bit and 64-bit HDR Lighting
    - Fully dynamic lighting
    - Intensive particle effects
    - Dynamic reflections
    - Refraction
    - Normal mapping

  • Let’s run the benchmark and see what it looks like ->
    • next
    Comment from Rutar @ 2007/05/16
    It's a about time someone takes on the noobs of Futuremark.
    Comment from Wolf2000me @ 2007/05/16
    Originally Posted by Rutar View Post
    It's a about time someone takes on the noobs of Futuremark.
    Absolutely. I couldn't agree more.