即将下水的国产航母与辽宁号相比有哪些改进之处

2X the speed of the previous generation for single-precision floating-point (FP32) operations provides significant performance improvements for graphics and simulation workflows on the desktop, such as complex 3D computer-aided design (CAD) and computer-aided engineering (CAE).

With up to 2X the throughput over the previous generation, third-generation RT Cores deliver massive speedups for workloads like photorealistic rendering of movie content, architectural design evaluations, and virtual prototyping of product designs. This technology also accelerates the rendering of ray-traced motion blur with greater visual accuracy.

Fourth-generation Tensor Cores, using the FP8 data format, deliver more than 2X the AI performance of the previous generation, enabling rapid model training and inferencing directly on RTX-powered AI workstations.

With 48GB GDDR6 memory, RTX 6000 gives data scientists, engineers, and creative professionals the large memory needed to work with large datasets and workloads like rendering, data science, and simulation.

Eighth-generation dedicated hardware encoder (NVENC) with AV1 encoding unlocks new opportunities for streamers, broadcasters, and video conferencing. It’s 40% more efficient than H.264, allowing users streaming at 1080p to increase their resolution to 1440p while running at the same bit rate and quality.

Support for NVIDIA RTX Virtual Workstation (vWS) software allows a personal workstation to be repurposed into multiple high-performance virtual workstation instances, letting remote users share resources to drive high-end design, AI, and compute workloads.