COPY from STDIN vs file with large CSVs

I have a CSV that's ~30GB. Some 400m rows. Would there be a meaningful performance difference to run COPY from STDIN using: cat f.csv | psql "COPY .. FROM STDIN WITH CSV" versus just doing "COPY ... FROM 'f.csv' WITH CSV"?

Thanks. It took about four hours to copy one and I felt that was a little much.

--