Amazon S3 access control + bandwidth

I'm thinking about maybe using Amazon S3, but have a few questions:

• does it provide access control to limit GETs?
• does it provide "high-speed" transfer?

I'm looking to serve large files to a vast number of end users. Each end user need to authenticate himself to our service before getting the file. Is this something that Amazon S3 can do? (Ive seen no scripting features at S3). I've seen a lot of authentication stuff and access control stuff, but it only seems to pertain to developers authenticating themselves, not the end user doing GETs. I'm hoping/thinking Amazon S3 would be a fast alternative (when talking bandwidth troughput) than just serving files through EC2, which turned out to be completely unsatisfying.

Solution 1:

"high-speed" transfer => CloudFront

Access control => not via S3, but you can use "long" URLs as workaround, such as http://aws.../uzuz967678678ghhhj/yourfile.pdf