Given the high volume of read requests and the need for low-latency access on a global scale, Amazon DynamoDB with Global Tables would be the optimal choice. DynamoDB is a managed NoSQL database service that can handle high-velocity read and write workloads and is designed for scalability and performance. Global Tables provides multi-region, fully-replicated databases, which would allow for fast, local reads and writes to take place near the user base, further decreasing latency and improving the user experience. Amazon Aurora would be an appropriate choice for high-throughput reads and writes if the global distribution and minimal latency were not as critical. Amazon Redshift is optimized for complex querying operations and large scale data warehousing, and thus less suitable for the described use case. Amazon RDS provides relational database service suitable for a variety of applications but does not inherently provide the same low-latency read performance at a global scale without significant additional configuration.