A mobile gaming company needs to capture data from its game app and make it available to three internal consumers. Each record is about 20 KB. The company wants optimal throughput from each device and requires a stream-processing application that provides dedicated throughput per internal consumer. Which solution meets these requirements?

The correct solution uses Amazon Kinesis Data Streams with the PutRecords API and enhanced fan-out. PutRecords efficiently sends multiple records in a single request, optimizing throughput from the mobile app. Kinesis Data Streams is designed for real-time stream processing. Enhanced fan-out provides dedicated throughput for each consumer by delivering data directly from the stream, preventing contention and ensuring each internal consumer receives data at its own rate without impacting others. Incorrect options: Kinesis Data Firehose is primarily for loading streaming data into data stores, not for dedicated throughput stream processing for multiple consumers. There is no "PutRecordBatch" API for Firehose, and dedicated throughput isn't enabled via a support case in this manner. Kinesis Producer Library (KPL) is for Kinesis Data Streams, not Firehose. Enhanced fan-out is a Kinesis Data Streams feature. While Kinesis Data Streams is correct, hosting stream-processing applications on EC2 instances without enhanced fan-out would mean all consumers share the same stream throughput, leading to potential contention and not meeting the "dedicated throughput per internal consumer" requirement.