• Confluence is currently running with all non-system plugins disabled, some functionality may be missing
Child pages
  • System Specifications FAQ
Skip to end of metadata
Go to start of metadata

1. Which is the best hardware configuration for using Maxwell?

 Click here to expand...

For the CPU engine (Draft and Production) each hardware component is in fact crucial in a particular part of the 3D process:

  • The CPU is the most important component for fast renders. Get the fastest processors you can afford, plus with network rendering, Maxwells render speed scales almost linearly when adding more computers to contribute to the render process. At least a quad-core CPU is recommended.
  • RAM is where the scene information is stored during the render process. If your scene has complex geometry, huge textures, set to render at high resolution, or has the MultiLight feature enabled (which increases the amount of RAM needed to store all the emitters information separately) then you may need a computer with a lot of RAM (4GB and more). A usual configuration for a rendering / video compositing computer is 12-16 GB of RAM.
  • The graphics card is not involved in the rendering process when using these engines. It is only involved in the openGL camera navigation when you are creating your scenes. A gaming card will be sufficient for most tasks, unless you plan to work with scenes containing many thousands of objects and/or require antialiased viewports which are much better handled by professional graphics cards. Their added advantage is that their drivers have been certified to work properly with different CAD / 3D applications such as SolidWorks, Rhino, Maya, 3dMax etc. It is also recommended to have a card with large on-board memory (1GB or more) if you plan to work with many large resolution textures and want to view them in theopenGL viewports.

For the GPU engine:

  • A graphics card based on CUDA is mandatory for the moment. We are supporting only Maxwell and Pascal architectures (check the table in this link for reference). Kepler architecture could work but the experience won't be good and probably the CPU engine will beat it on speed.
  • The memory of the graphics is crucial as the whole scene has to fit in it to be able to render, so the higher the better.
  • The higher the number of CUDA cores and their speed, the better. This will determine the rendering speed for that particular graphics card.
  • For the moment Maxwell can only render in one GPU per machine but in the near future, we will add support for multi-GPU.

For the Denoiser:

  • Denoiser is capable of running on CPU or GPU, with NvidiaAMD and Intel graphics cards; it works with CUDA (Nvidia) and also can work using OpenCL either in GPU (Nvidia, AMD or Intel) or CPU.
  • When using GPU, we recommend having a graphics card with at least 2.5GB. This number is actually dependent on frame size, 2.5GB is for 2k square imagery.
  • When using Nvidia CUDA, the drivers should support at least CUDA 7.0
  • When using AMD, the Catalyst driver should be up to date.
  • In the case you get an error message when using CPU about OpenCL ICDs, it will probably be solved by installing one of these packages depending on your graphics card:
  • The Denoiser will first try to use CUDA, if it doesn't find a compatible card, it will fallback to OpenCL GPU (usually with AMD and Intel cards). If the image doesn't fit in the graphics card memory, you can then use OpenCL with CPU (which will use RAM memory).

2. What are the minimum system requirements? 

 Click here to expand...

The minimum system requirements for Maxwell Render are listed in the System requirements page.

 3. Which parameters determine the amount of RAM needed for certain project? 

 Click here to expand

The amount of RAM needed depends on several factors:

  • Resolution: The bigger the render the more render calculations Maxwell has to keep in RAM.
  • Multilight: Each separate light sliders needs to be stored in RAM so the more separate sliders you have, the more RAM is needed. To minimize the RAM usage you can apply the same emitter material to several geometry objects that are ment to be emitters in which case they will all use the same slider - saving RAM. Color ML will need more RAM than regular ML because each color channel per light also needs to be stored in RAM.
  • Size of textures: Every texture used in the scene, including IBL maps, need to be stored in RAM.
  • Geometry: Although the amount of triangles plays a smaller part in the RAM usage compared to the points above, it still affects the RAM needed but in most scenes this will be of minor importance.
  • Pretesselated displacement: This type of displacement subdivides the geometry before rendering (unlike On-the-Fly displacement), which renders fast but needs to hold the extra geometry in RAM. The higher the subdivision setting in the displacement material, the more RAM is needed.
  • Extensions: Hair and particles need more RAM - the more segments the hair has, or the more particles you have. Although these extensions are very efficient, they will use up RAM if used in the scene.
  • Extra render channels: The extra channels you can specify in the render options (Alpha, MatID, ObjectID, Normals etc) will need extra RAM to render. The amount they need depends also on the resolution of the main render.


Taking into account all these factors it is impossible to say how much RAM you will need for a certain render. It is best to test a render on your machine and check the RAM usage during the render. On Windows 32 bit computers, each application can use a maximum of 1.6GB of RAM (or 3GB by changing the boot.ini file - see the next FAQ). This will be limiting if you want to do higher resolution renders. It is strongly suggested to use a 64 bit system which can use up to 128GB of RAM.


Related content:
Error formatting macro: contentbylabel: com.atlassian.confluence.api.service.exceptions.BadRequestException: Could not parse cql : null