NVIDIA and ComfyUI Enhance Local AI Video Generation for Game Developers at GDC
Article Content
“Our goal is to make AI video generation more accessible and efficient for creators,” stated NVIDIA representatives at the Game Developers Conference (GDC) in San Francisco. This week, NVIDIA unveiled a series of updates aimed at streamlining AI-driven video generation workflows on RTX GPUs and the DGX Spark desktop supercomputer.
Key announcements include the launch of ComfyUI’s new App View, which features a simplified interface designed to lower the entry barrier for users unfamiliar with complex node graphs. This mode allows users to input prompts, adjust basic parameters, and generate content with ease, while still retaining the option to switch back to the full node-based Node View. The update also introduces support for NVFP4 and FP8 data formats, delivering up to 2.5x faster performance and reducing memory usage by 60% on NVIDIA GeForce RTX 50 Series GPUs.
In addition, NVIDIA introduced RTX Video Super Resolution for ComfyUI, a real-time 4K upscaler that accelerates video generation workflows. Available as a standalone node and a Python package, this technology leverages RTX GPU Tensor Cores to upscale videos 30 times faster than other local upscalers while significantly lowering VRAM consumption. Developers can access sample code and documentation via GitHub and the PyPI repository.
Performance benchmarks were shared for generative AI models LTX-2 and FLUX.2 Klein 9B running on an NVIDIA RTX 5090 GPU, highlighting improvements in resolution, frame count, and processing steps. NVIDIA also promoted upcoming events such as the NVIDIA Studio Sessions tutorial and the GTC conference, which will focus on generative AI workflows for images, video, 3D, and PBR materials.
Additional tools and collaborations were noted, including the LTX Desktop open-source video editor optimized for NVIDIA GPUs, LM Link for remote AI model execution, and partnerships with Topaz Labs and Microsoft to enhance AI performance on consumer hardware. NVIDIA plans to release further updates to DLSS technology and RTX Remix modding tools in the near future.
Read more: blogs.nvidia.com