NVIDIA Launches New GPUs And Services for Generative AI Inferencing

The star of the show is a new PCI-card with 12 times more inference throughput for large models like ChatGPT.