Stable Video Diffusion's latest update allows older computers to run image-generated video models!

In last Tuesday's Stability AIhas released the latest version of its Stable Video Diffusion model, a powerful technology capable of converting images into video. The update also adds support for Stability AI's ComfyUI tool, a graphical user interface designed to help users build complex workflows using a graphical/node interface.

According to Stability AI, this update allows users to use a 7-year-old NVIDIA GTX 1080 graphics card (released in May 2016) to generate videos with a resolution of 1024*576 and a length of 25 fps. This is certainly good news for those who have older hardware but still wish to try this technology.

As for AMD users, they will need at least a 6800 XT graphics card to run ROCm in Linux and use this technology to convert images to video. However, this also means that this technology may have some limitations for those with AMD hardware.

In its blog post, Stability AI shared a sample flow of calling Stable Video Diffusion in ComfyUI for a 14 fps model and a 25 fps model. At this time, the technology is still in what Stability AI calls a "research preview" phase, which means that users who want to try the technology must agree to some terms of use. These terms define intended scenarios for Stable Video Diffusion (e.g., "educational or creative tools," "design and other artistic processes," etc.) and unintended scenarios (e.g., "the use of Stable Video Diffusion for "Factual or realistic representations of people or events").

Demo Pictures

The update to Stable Video Diffusion provides more options and flexibility for those looking to convert images to video. Whether users are using NVIDIA or AMD hardware, they can choose the right model and tools for their needs. At the same time, Stability AI's terms of use ensure that the technology is used appropriately and in the right application scenarios.