Content-aware lossless compression and decompression of floating point data

A method for compressing at least one floating point number, said method comprising:

• obtaining said at least one floating point number represented using one or more bits to indicate a sign of said at least one floating point number and one or more additional exponent bits to indicate an exponent at a given base and a significand of said at least one floating point number;
• aligning said at least one floating point number by moving a radix point of said at least one floating point number and updating said exponent so that said at least one floating point number is separated into an integral portion and a fractional portion, wherein a plurality of said fractional portions of said at least one floating point number comprises at least one repeating pattern;
• encoding said integral portion separately;
• and
• encoding said fractional portion of said at least one floating point number using a dictionary-based compression method.

A system for compressing at least one floating point number, comprising:

• a memory;
• and
• at least one hardware device, coupled to the memory, operative to implement the following steps:
• obtaining said at least one floating point number represented using one or more bits to indicate a sign of said at least one floating point number and one or more additional exponent bits to indicate an exponent at a given base and a significand of said at least one floating point number;
• determining a length of said significand equal to a number of bits between a most significant bit of said significand and a least significant bit of said significand having a predefined binary value;
• and
• encoding said at least one floating point number by encoding, as a single code, an exponent/length pair comprising said exponent and said length, wherein one or more significant bits following said least significant bit of said significand having said predefined binary value are not included in said encoding.

A system for compressing at least one floating point number, comprising:

• a memory;
• and
• at least one hardware device, coupled to the memory, operative to implement the following steps:
• obtaining said at least one floating point number represented using one or more bits to indicate a sign of said at least one floating point number and one or more additional exponent bits to indicate an exponent at a given base and a significand of said at least one floating point number;
• aligning said at least one floating point number by moving a radix point of said at least one floating point number and updating said exponent so that said at least one floating point number is separated into an integral portion and a fractional portion, wherein a plurality of said fractional portions of said at least one floating point number comprises at least one repeating pattern;
• encoding said integral portion separately;
• and
• encoding said fractional portion of said at least one floating point number using a dictionary-based compression method.