An online gaming company is migrating user data to Amazon DynamoDB to support growth. Current tables store user profiles, achievements, and in-game transactions. The company needs a resilient, highly available DynamoDB architecture to ensure a seamless gaming experience. Which solution meets these requirements MOST cost-effectively?

The correct answer is to use DynamoDB global tables for automatic multi-Region replication, deploy tables in multiple AWS Regions, use provisioned capacity mode, and enable auto scaling. This approach provides a resilient, highly available architecture by automatically replicating data across specified Regions, ensuring data availability even during regional outages. Provisioned capacity with auto scaling is cost-effective for predictable workloads, as gaming often has, allowing for optimized cost while maintaining performance. Creating tables in a single Region and using global tables to replicate data across multiple Regions is redundant; global tables inherently involve multiple Regions. Using DAX and manual Cross-Region Replication is less efficient for high availability than global tables. Creating tables in multiple Regions and using DynamoDB Streams for Cross-Region Replication is possible but requires more manual configuration and management than global tables, which automate this process.