Google Cloud A4X Max Bare Metal Instances Support 50k GPU Clusters, Network Bandwidth Doubles
Google Cloud announces that the A4X Max bare-metal instances can scale up to a 50k GPU cluster, with network bandwidth doubled compared to the previous generation. The series covers A4X/A4/A3/A2/G4/G2, providing recommendations based on workloads such as pre-training, fine-tuning, inference, graphics, and HPC; pricing is based on pre-installed GPUs, vCPUs, memory, and local SSDs, supporting on-demand, Spot, Flex-start, reserved, and other billing options, with differences in maintenance experience.