You work for an online grocery store. You recently developed a custom ML model that recommends a recipe when a user arrives at the website. You chose the machine type on the Vertex AI endpoint to opti

Maintaining the same single machine without autoscaling will cause the endpoint to be overwhelmed by a 4x traffic increase, leading to performance issues or outages.

Changing to a larger machine type (32 vCPUs) increases individual machine capacity but does not provide dynamic scaling for fluctuating traffic and would lead to over-provisioning and higher costs during off-peak times.

Configuring autoscaling on the Vertex AI endpoint allows the system to automatically provision more instances of the currently cost-optimized machine type to handle increased traffic during peak periods. This ensures efficient resource utilization and cost management by scaling up during high demand and scaling down when traffic subsides.

Changing to a GPU machine type is an unnecessary and costly change as the model was already optimized for and running on a CPU, and GPUs are typically reserved for specific types of intensive deep learning workloads.