Eliminating the “barrier of language” in the world and realizing free interaction between people on a global level will require the development of highly sophisticated algorithms and massive volumes of speech and text data.ASTREC uses high-speed GPU computers, large-scale memory machines, and cluster servers to process these massive volumes of data and leverage them in the development of multilingual speech translation technology.