A company regularly uploads GB-sized files to Amazon S3. After upload, a fleet of Amazon EC2 Spot Instances transcodes the files. The company needs to scale throughput when transferring data from the on-premises data center to Amazon S3 and when downloading data from Amazon S3 to the EC2 instances. Which solutions will meet these requirements? (Choose two.)

S3 multipart uploads improve throughput for large files by splitting them into smaller parts that can be uploaded concurrently. This significantly reduces the total upload time, which is crucial for GB-sized files. Fetching multiple byte-ranges of an object in parallel (also known as S3 Select or Range GETs) allows applications to download specific parts of an object simultaneously. This parallelization increases download throughput, especially for large files being processed by EC2 instances. Using S3 bucket access points does not directly improve data transfer throughput; they provide a way to manage access to S3 buckets. Uploading files into multiple S3 buckets or adding a random prefix to each object primarily addresses S3 request rate performance for high-volume access patterns, not the throughput of individual large file transfers.