Scaling AI Inference Across Multiple GPUs Using NVIDIA TensorRT with Multi-Device Inference Support
产业动态
来源：NVIDIA 技术博客发布时间待核实
Generative AI workloads are rapidly outgrowing the memory and compute budget of single GPUs. For inference developers building media generation pipelines, the... Generative AI workloads are rapidly outgrowing the memory and compute budget of single GPUs.