Big Data Challenges of FAST

Abstract

We present the big-data challenges posed by the science operation of the Five-hundred-meter Aperture Spherical radio Telescope (FAST). Unlike the common usage of the word `big-data', which tend to emphasize both quantity and diversity, the main characteristics of FAST data stream is its single-source data rate at more than 6 GB/s and the resulting data volume at about 20 PB per year. We describe here the main culprit of such a high data rate and large volume, namely pulsar search, and our solution.