How big is the latency between these edge locations and AWS and Google GCP regions? Since if the DB is hosted on AWS or GCP, then there would be latency going

How big is the latency between these edge locations and AWS and Google GCP regions? Since if the DB is hosted on AWS or GCP, then there would be latency going to that too. That's what I'm really curious about. Vercel being mainly on AWS would make the latency to the DB optimal if it is hosted on AWS in the same region. But what if the DB is on GCP?