| 摘要 |
A method is disclosed for compressing binary numbers to be stored in a memory. A first number to be stored is first determined to be in one of several bit size ranges, and depending on the bit size range a fixed value number is subtracted from the first number. A second number resulting from the subtraction process has two binary bits are prefixed thereto and the combination is stored in memory. The two binary bit prefix identifies the particular bit size range and thereby the fixed value number originally subtracted from the first number. On reading out the second number with two bit prefix, the fixed value number identified by the two bit prefix is added back to the second number to get the first number. The combination of the second number plus two bit prefix has fewer total bits than the first number thus achieving compression of the first number.
|
