Network Rendering was first introduced with OctaneRender 2. Texture count limitations Compute CapabilityĦ4 LDR (8 bit) RGBA, 32 LDR (8 bit) greyscale, 4 HDR (16 bit) RGBA and 4 HDR (16 bit) greyscaleġ44 LDR (8 bit) RGBA, 68 LDR (8 bit) greyscale, 10 HDR (16 bit) RGBA and 10 HDR (16 bit) greyscale If there’s not enough VRAM, then you can enable OctaneRender’s Out-Of-Core features.Īs OctaneRender does not use the CPU for rendering, a fast multi-core CPU is not required, but it does significantly improve scene-loading speeds.Ĭompute capabilities required for octane features: Compute CapabilityĦ.0 and 6.1 (eg: Nvidia Titan X, 1080, 1070) Therefore, this function is limited by the card with the least amount of VRAM. All textures, HDRI, geometry, and other scene elements must fit into a functional memory – each render GPU must have the same copy of all of the scene elements to complete the process. NOTE: Using multiple GPUs together combines the cores, resulting in a linear speedup, but VRAM is not compounded. The official list of NVIDIA CUDA-enabled products is located at. OctaneRender scales perfectly in a multi-GPU configuration and can use different types of NVIDIA cards at once - e.g., a GeForce RTX 2080Ti combined with a Quadro RTX 6000. GeForce cards are fast and cost-effective, but have less VRAM than Quadro and Tesla cards. GPUs from the GeForce® line are usually clocked higher and render faster than the more expensive Quadro® and Tesla GPUs. Texture limits and differing power efficiency ratings also apply depending on the GPU microarchitecture.

OctaneRender also requires a minimum of 8 GB RAM, and we recommend 16 GB or more.

RTX™ 20 Series, Titan RTX, GeForce RTX, and GTX 16 Series), Quadro® GPUs, Volta™ GPUs, GTX Titan (high-end), Pascal™ (GTX 10xx), Maxwell (GTX 7xx, GTX8xx, GTX9xx), and Kepler™ (GTX 680, GTX 690).

OctaneRender requires the latest CUDA® 10 drivers and a CUDA-enabled NVIDIA® video card with support for compute capability 3.0 or higher.

Opengl 4.1 renderer not supported