Skip to main content

CINEBENCH

Technical Information

  • Cinebench 2026 tests GPU and CPU performance by using Maxon’s powerful Redshift render engine.
     
  • Cinebench 2026 supports AMD and Intel x86/64 CPUs, Apple’s M1, M2, M3, M4 & M5-powered computing systems, the Snapdragon compute platform and Ampere Altra processors.
     
  • Cinebench 2026 supports GPU accelerated Metal RT hardware ray tracing starting with M3-powered Macs (no changes for previously supported Apple silicon hardware).
     
  • Cinebench 2026 adds support for NVIDIA 5000 series GPUs and the AMD Radeon 9000 series, as well as for NVIDIA Hopper and Blackwell data center GPUs.
     
  • Cinebench 2026 adds a test to evaluate the performance of SMT enabled cpu cores.
     
  • Cinebench is based on the latest Cinema 4D 2026 and Redshift code using updated compilers (Clang V19) and has a minimum runtime activated by default.
     
  • Cinebench 2026 provides improved benchmark accuracy for current and next generation CPUs to test if a machine runs stable on a high CPU load, if the cooling solution of a desktop or notebook is sufficient for longer running tasks to deliver the full potential of the CPU and if a machine can handle demanding real-life 3D tasks.
     
  • Users have the option to directly test the single core performance without manually enabling the “Advanced benchmark” option. The “Advanced benchmark” allows users to set arbitrary minimum runtimes to stress test the hardware for even longer periods of time.
     
  • Because of the code and compiler changes that have accelerated the rendering of the scene, Cinebench 2026 score values are readjusted to a new range so they should not be compared to scores from previous versions of Cinebench.
     
  • Compared to Cinebench R23 the computational effort for the scene increased sixfold in the multithreaded rendering test. This reflects the CPU performance improvements as well as the increased demands artist have to deal with these days.
     
  • Depending on your CPU configuration Cinebench 2026 will need at least 6.5 - 8.5 GB RAM. The minimum memory requirement is therefore set to 16 GB (while macOS can execute the CPU test also on 8 GB machines – with significant influence of paging – Windows usually has several GB assigned to non-purgeable memory which prevents the execution of Cinebench 2024 on 8 GB machines).
     
  • GPUs need more memory for efficient operation, and as such, they need 8 GB or more of GPU memory. On Apple silicon, this memory is part of the unified memory system, and therefore Redshift GPU can only run on machines with at least 16 GB.
     
  • Cinebench 2026 will not execute on unsupported processors.
     
  • On systems lacking sufficient RAM to load the test scene, a warning will be displayed, and the CPU and/or GPU benchmark will not be executed.
     
  • Background tasks can significantly influence measurement and create diverse results. It's always a good idea to shut down all running programs and disable any virus checking or disk indexing but it's impossible to eliminate all background processes. Modern operating systems perform various background tasks that cannot or should not be disabled, even though they could have a minor influence on the results.
     
  • Test results can vary slightly because it's impossible to disable every background task of an operating system. These tasks are one factor that may have a slight influence on measurements. Also, modern computers and graphics cards dynamically adjust clock speeds based on environmental conditions like power and temperature. For instance, processors will reduce clock speed when running too hot to allow for cooling and prevent damage. With many modern processors, the reverse is also true. They can overclock themselves when the temperature is low enough. Therefore, a system freshly started in a relatively cool environment will typically run faster than the same system that has been performing benchmarks for several hours in a heated office.
     
  • It is also possible to launch Maxon Cinebench with command line options. Please refer to your operating system manual on how to start an application using the command line. After the name of the application enter one of the options listed below. Maxon Cinebench will then be executed, run the specified test, then quit and display the result in the command line console. The result is not saved as a file.
     
  • The following command line options are available:
    • g_CinebenchCpu1Test=true – runs only the Single Thread test procedure
    • g_CinebenchCpuSMTTest=true – runs only the Single Core test procedure on a SMT capable core
    • g_CinebenchCpuXTest=true – runs only the Multiple Threads test procedure
    • g_CinebenchAllTests=true – runs all test procedures sequentially
    • g_CinebenchMinimumTestDuration=100 – sets a minimum test duration of 100 seconds
       
  • To get a proper console log on Windows, you must add an additional command before the Cinebench executable name. Example:
    • start /b /wait "parentconsole" Cinebench.exe g_CinebenchAllTests=true

Limitations

There is no "limitless" or "without compromises" in real life and engineering:

  • The range of hardware covered on GPU and CPU is vast. For extreme GPU configurations (e.g. having half a dozen or more of the most powerful GPUs in one machine) Maxon’s Redshift GPU benchmark creates a more demanding workload. 
    The Redshift GPU benchmark can keep these hardware setups busy for several minutes with its benchmark scene (on the flip side this will get quite slow on non-high end GPUs and you can’t compare those results relative to the CPU in your machine anymore)
     
  • Cinebench unfortunately can’t measure the wall power to determine how much energy was used for the render task. Depending on your use case - e.g. working on a notebook (without having access to power connector) or having high energy costs do deal with - energy efficiency of the hardware vs render speed might be an important metric for you to consider.

System Requirements

Operating Systems

  • Windows 10 Version 20H2 or higher or Windows 11 for x86/64 hardware
  • Windows 11 for arm64 hardware
  • Apple macOS 14.7+ (Sonoma)

Minimum Requirements (Windows)

  • 16 GB of RAM
  • 64-bit Intel or AMD cpu with AVX2 support or Windows 11 on Snapdragon compute platform or ARM v8.1 64-bit CPU
  • NVIDIA GPU with CUDA compute capability 5.0 or higher and 8 GB VRAM, or
  • AMD "Navi" or "Vega" GPU or later with HIP capability and 8 GB VRAM or more (see GPU list below)

Minimum Requirements (macOS/Intel)

  • 16 GB of RAM
  • 64-bit processor with SSE4.2 support
  • AMD "Navi" or "Vega" GPU or later with 8 GB VRAM or more (see GPU list below)

Minimum Requirements (macOS/Apple Silicon)

  • 16 GB of unified memory; CPU rendering works on Apple Silicon machines with 8 or 12 GB of memory too, but OS memory paging can degrade the performance compared to machines with 16 GB of memory
  • Apple “M” Series (M1, M1 Pro, M1 Max, M1 Ultra, M2, M2 Pro, M2 Max, M2 Ultra, M3, M3 Pro, M3 Max, M3 Ultra, M4, M4 Pro, M4 Max, M5)

Supported AMD GPUs for Windows:

  • Radeon 9000 series
  • Radeon PRO W7900, W7800, W6800, W6600, VII and W5700
  • Radeon RX 7950XTX, RX7950XT, RX7900, RX6950, RX6700XT, RX6600 and RX5700XT

In addition the following GPUs should be supported (but are not actively tested):

  • Radeon Vega Frontier Edition, Radeon RX Vega 56, Radeon RX Vega 64 (gfx900)
  • Radeon RX 5500 (gfx1012)
  • Radeon RX 6800, Radeon RX 6900 (gfx1030)
  • Radeon RX 6700 (gfx1031)
  • Radeon RX 7800 (gfx1101)
  • Radeon RX 7600, Radeon 7700 (gfx1102)

Supported AMD GPUs for macOS/Intel:

  • None