When hardware encryption is properly implemented, each encrypted drive ships with a unique key. Each sector is transparently encrypted and decrypted with this key during write and read operations. The key is stored in a hidden sector within a System Area reserved for the firmware. If this key is lost or intentionally deleted, the data become gibberish. This means that an encrypted drive can be securely erased in a split second simply by throwing away all copies of the original key and then generating a new key. This is how cryptoerase works and is particularly advantageous for SSDs. The alternative for ordinary HDDs and SSDs is to zero-fill every sector. In the case of SSDs, this would consume another P/E cycle, thus increasing wear and tear.
To protect one's data on an ordinary HDD or SSD requires setting a password which then needs to be stored somewhere on the drive. If this password is defeated, the non-encrypted data are immediately accessible. An encrypted HDD or SSD, OTOH, uses the password to encrypt the key, and this encrypted key then replaces the original key. Ideally the original key should be discarded. This means that the user must supply the correct password in order for the drive to be able to regenerate the original key from the encrypted one.
Therefore the advantages of a Self Encrypting Drive (SED) are :
1/ Instant cryptoerase by deleting the key
2/ The key is encrypted by the password
3/ Neither the password nor the original key are retained by the drive
4/ The correct password is required to decrypt the key