FPcrush Real-Time Floating-Point Compressor Generator v1.0
FPcrush generates a custom compression algorithm for each input file and uses this algorithm to compress the data. It is currently available on the Maverick supercomputer at TACC. FPcrush's operation is described in this SC paper.
To compress a file that contains double-precision IEEE 754 floating-point data, run:
/home/00976/burtsche/FPcrush/FPcrush_double input_file compressed_file
To decompress the compressed file, run:
/home/00976/burtsche/FPcrush/FPuncrush_double compressed_file output_file
For single-precision files, replace the "double" with "single" in the above commands.
Sample single-precision data sets are available here
and double-precision data sets are available here.
This work has been supported in part by the National Science Foundation under Grant No. 1438963.