The Nvidia GeForce RTX 5070 TI graphics card (built on the Nvidia Blackwell Architecture) is ready to accelerate the creation of generative AI content and creative performance.
The GeForce RTX 5070 Ti GPU features a fifth generation tensor core that supports FP4, doubles performance and reduces VRAM requirements for running generated AI models.
Additionally, the GPU is equipped with two 9th generation encoders and a 6th generation decoder that adds support for a 4:2:2 pro-grade color format, and enhances the encoding quality of HEVC and AV1. This combo accelerates your video editing workflow and reduces export time by 8x compared to a single encoder GPU that doesn’t have 4:2:2 support like the GeForce RTX 3090.
The GeForce RTX 5070 Ti GPU also includes 16GB of high-speed GDDR7 memory and 896GB/s total memory bandwidth.
NVIDIA DLSS 4 is a set of neural rendering technologies that use AI to increase frames per second (FPS) and improve image quality, and is now available in professional 3D apps such as Chaos Vantage. D5 Render adds DLSS 4 to the beta version with new multi-frame generation capabilities to triple frame rates. 3D rendering software Maxon Redshift also added support for Nvidia Blackwell, resulting in a 30% increase in performance.
The February Nvidia Studio driver with support from GeForce RTX 5070 Ti GPU is ready to download next week. For automatic studio driver notifications, please download the NVIDIA app.
Use NVIDIA’s product finder to pick up your GeForce RTX 5070 TI GPU or pre-built systems. Retail partners list the models available, so please return regularly after 6am PT. Check the complete specifications.
Ready for the generation AI era
Black Forest Lab’s suite of FP4-OPTIMIZED FLUX.1 (DEV) image generation models are now available for embracing faces.
FP4 is a low quantization method that reduces the model size, similar to file compression. Flux.1 (DEV) on FP4 requires less than 10GB of VRAM, compared to over 23GB on FP16.
This means that cutting-edge Flux.1 (DEV) models can run on GeForce RTX 5070 Ti GPUs and all GeForce RTX 50 Series GPUs. This is important. This is because the Flux.1 (DEV) model cannot be run on FP16 with memory constraints.
With the GeForce RTX 5070 Ti GPU, the Flux.1 (DEV) model can generate images in just 8 seconds on the FP4 compared to 20 seconds on the FP8 on the GeForce RTX 4070 Ti GPU.
A versatile viewport
DLSS 4 is now available for D5 rendering in Chaos Vantage and Beta.
https://www.youtube.com/watch?v=QQN3BSPNTYI
Both apps natively support the improved super resolution and ray reconstruction model of DLSS 4 (with transformers), increasing image detail and improving stability.
D5 Render supports DLSS multiframe generation in DLSS 4 to increase frame rates by using AI to generate up to three frames per traditional rendered frame.
This allows animators to smoothly navigate the scene with framerate multiplication, allowing them to render 3D content at 60 fps or more, even at large file sizes.
Maxon Redshift, a 3D rendering software that uses GPU acceleration to visualize 3D models, scenes, animations and designs, brings updates to fully utilize the GeForce RTX 50 series GPUs and accelerates performance by up to 30%. It’s been released.
It brings new creative app updates and optimizations each month with the NVIDIA Studio platform. Follow Nvidia Studio on Instagram, X and Facebook. Get updates directly in your inbox by accessing the tutorial on your Studio YouTube channel and subscribing to Studio Newsletter.
Please refer to the Software Product Information Notice.
