características y beneficios
- Double-speed processing for single-precision floating point (FP32) operations
- Up to 2X throughput over the previous generation with second-generation RT Cores
- New Tensor Float 32 (TF32) precision provides up to 5X training throughput over the previous generation
- Increased GPU-to-GPU interconnect bandwidth with Third-Generation NVIDIA NVLink
- Ultra-fast GDDR6 memory, scalable up to 96 GB with NVLink
- Ideal for CAD, CAE, photorealistic rendering, AI, data science, and simulation