Second generation RT cores
With twice the throughput of the previous generation and the ability to simultaneously run ray tracing with shading or noise reduction functions, second-generation RT cores deliver significant acceleration for workloads such as photorealistic movie content rendering, architectural design estimating, and virtual prototyping. product design. This technology also speeds up the rendering of ray-traced blur, delivering faster results and greater visual fidelity.
Third generation tensor cores
The new precision Tensor Float 32 (TF32) delivers up to 5x the training performance of the previous generation, accelerating the training of AI and data science models without any code changes. Hardware support for structural sparsity doubles inference throughput. Tensor Cores also bring AI to graphics with capabilities such as DLSS, AI denoising, and improved editing for select applications.
48 GB GPU memory
Ultra-fast GDDR6 memory, scalable up to 96GB with NVLink, gives scientists, engineers and creative professionals the massive memory they need to handle large data sets and workloads such as data science and simulation.
Next-generation enhancements through NVIDIA Virtual GPU (vGPU) software enable larger, more powerful virtual workstation instances for remote users, enabling high-quality remote design, artificial intelligence, and compute workloads.