The kana alphabets were (probably) borrowed from Pali, and syllables follow the structure (consonant + vowel) or (vowel). In other words, a consonant must not, grammatically, occur alone. I don’t know if Japanese still retains this as an explicit rule, but this is why you see the -u ending. It may or may not be pronounced, depending on which way flows better.
Many other languages with Pali / Sanskrit heritage have similar behaviour. However, Sanskrit itself and some modern languages have a dedicated character called a ‘viraama’, which says ‘this consonant has no following vowel’. For example, in the word ‘Padma’ (lotus), the d is followed by a virama. Other languages, like Japanese, use ‘u’ instead of a dedicated viraama. So different languages in east, southeast and south Asia might write and say it as Padma, Paduma or Padama, but all versions would be mutually intelligible.