Game developers and artists are building cinematic worlds and iconic characters, raising the bar for immersive gaming experiences. NVIDIA RTX AI PC.
At the Game Developers Conference (GDC) in San Francisco this week, NVIDIA announced a series of updates that streamline AI video generation for concept development and storyboarding on RTX GPUs and NVIDIA DGX Spark desktop supercomputers.
These announcements include:
- ComfyUI’s new app view features a simplified interface, lowering the barrier to entry for popular generative AI tools.
- RTX video super resolution is available in ComfyUI, the perfect real-time 4K upscaler for video generation. It is also available to developers as a Python Wheel.
- FLUX.2 Klein is currently available in NVFP4 and FP8 model variants, with NVFP4 support for LTX-2.3 coming soon, delivering up to 2.5x performance improvement and 60% memory usage reduction for both models.
Frictionless local AI: Collaborate, optimize, and customize
Many of today’s popular AI applications make it easy for beginners to try out cutting-edge models directly on their laptops or desktops.
For artists who aren’t familiar with node graphs, ComfyUI’s new app view presents workflows in a simplified interface. Users simply fill in the prompts, adjust some simple parameters, and press generate. The full node-based experience is still available as node view, allowing users to seamlessly switch between the two modes.
App View is compatible with ComfyUI’s RTX optimization. RTX GPU performance is 40% faster since September, and ComfyUI now natively supports NVFP4 and FP8 data formats. All combined, the NVFP4 format on NVIDIA GeForce RTX 50 Series GPUs delivers 2.5x faster performance and 60% less VRAM, and FP8 delivers 1.7x faster performance and 40% less VRAM.
At CES in January, NVIDIA announced several models released with support for NVFP4 and FP8. And now more NVFP4 and FP8 models are available — LTX-2.3, and Support for NVFP4 is coming soon. FLUX.2 Klein 4Band FLUX.2 Klein 9B — Directly in ComfyUI. First, download the NVFP4 and FP8 checkpoints directly from Hugging Face, load the default workflow into ComfyUI via the template browser, and replace the default model checkpoints with the newly downloaded checkpoints.
App view mode is currently available. Click here for details Comfortable UI.
Faster 4K video generation
Obtaining high-quality video output often requires juggling three constraints: speed, VRAM, and control. Many artists ultimately want 4K quality, but most prefer to generate a small, fast preview first and then upscale it. Today’s upscalers take minutes to upscale a 10 second clip to 4K resolution.
Users can now instantly upscale their generated videos to 4K using NVIDIA RTX video super resolutions available as nodes in ComfyUI. RTX Video can be accessed as a standalone node for building video workflows from scratch.
For AI developers, NVIDIA has released a free Python package. PyPI repositorywith Sample code on GitHub and VFX Python Binding Guideyou can get started right away. This package provides programmatic access to the same AI upscaling technology that powers RTX video and runs directly on RTX GPU Tensor Cores, delivering 4K upscaling 30x faster than other popular local upscalers and at a fraction of the VRAM cost. The package power supply is as follows: NVIDIA Video Effects Software Development Kit.
Ready to get started with ComfyUI? Check out the latest NVIDIA Studio Sessions tutorials hosted by visual effects artists. max novak For guided walkthroughs:
#ICYMI: Latest RTX AI PC updates at GDC
🎉How to join NVIDIA GTCIt will be held in San Jose from March 16th to 19th! check out “Create generative AI workflows for design and visualization with ComfyUI” On March 17, a training session led by NVIDIA 3D Workflow Specialists focused on building RTX accelerated production workflows for images, video, 3D, and PBR materials. register Today is session catalog.
💡LTX desktop is a fully local open source video editor that runs directly on the LTX engine and is optimized for NVIDIA GPUs and compatible hardware.
🦥 LM link Connect separate devices running LM Studio so you can run models on remote machines as if they were local. This is ideal for users who want to run agents on their laptops while accessing free private AI powered by DGX Spark or RTX desktop. learn how to run LM Studio on DGX Spark.
🎮On Tuesday, March 31st, NVIDIA DLSS 4.5 Dynamic Multi-Frame Generation and DLSS 4.5 Multi-Frame Generation 6x Mode Override will be released for GeForce RTX 50 Series owners as part of the next opt-in NVIDIA app beta. learn about NVIDIA News at GDC.
🤖Next month, New NVIDIA RTX Remix update introduces Advanced Particle VFXThis allows modders to create a wide range of particle effects that further improve image quality, detail, and immersion.
🦄topaz lab worked with NVIDIA to optimize NeuroStream for NVIDIA GPUs. This is a unique VRAM optimization that allows complex AI models to run on consumer hardware.
📃Microsoft introduced support for VoiceMod, one of the first apps to enable Windows ML for GPU inference, significantly improving performance voice quality compared to CPU.
Plug into your NVIDIA AI PC. facebook, Instagram, TikTok and × — Subscribe to stay up to date RTX AI PC Newsletter. Follow NVIDIA Workstation linkedin and ×.
look news About software product information.
