The directory on the server that receives the upload is not controllable and is staggered out to be able to hold that much (the computer ID is used to make multiple levels of directory hierarchy to allow the file system to hold things)
Once the files are uploaded and extracted most users of these files copy them away to where they need them to process them.
Note that the client will only upload changed files if the last archive was of the same content. So if you repeatedly upload the same directory for example it will only upload changed files but continue to list all the files that were in the directory, as it assumes that the server already has these files.