Data packing for power and area-efficient memory structures and performance-efficient decoding of encoded data

A method, comprising:

• receiving, by a controller, compressed data including data items and qualifier bits for the data items;
• discarding, by the controller, the data items designated as invalid by the qualifier bits;
• generating, by the controller, data type bits specifying data type information for the data items designated as valid by the qualifier bits;
• and
• storing the data items of the compressed data designated by the qualifier bits as valid and the data type bits within a buffer.

A method of decoding data by one or more decoders, the method comprising:

• processing, by a plurality of processing sections of each decoder, one or more bits of encoded data from a work buffer and generating a signal specifying whether the one or more bits of encoded data are a code or are part of a code;
• passing, by a switch, a data item output from a selected processing section of the plurality of processing sections, wherein the switch selects the selected processing section based on the signal generated by each processing section of the plurality of processing sections;
• passing the data item, by a de-mapper circuit of each decoder, in response to the data item being a literal, or outputting, by the de-mapper circuit, a length value or a distance value in response to the data item being a length or a distance;
• and
• selectively incrementing, by an adder circuit of each decoder, the length value or the distance value based on whether the data item has one or more extra bits and a value specified by the one or more extra bits.

A method of decoding data by a decoder, the method comprising:

• during each of a plurality of iterations, concurrently testing, by each processing section of a plurality of processing sections of the decoder, whether one or more bits of encoded data are a first code or are part of a code, wherein in each iteration, each processing section operates on a different set of the one or more bits of the encoded data and generates a signal specifying a result of the testing;
• for each iteration, generating an address of a memory of a selected processing section based on the one or more bits of encoded data and outputting a data item derived from the one or more bits of encoded data processed by the selected processing section;
• testing the data items for the plurality of iterations to determine whether the data item is at least one of a literal, a length, or a distance;
• for data items that are literals, outputting the literals;
• and
• for data items that are lengths or distances, outputting a length value or a distance value and selectively updating the length value or the distance value based on whether the data item is associated with one or more extra bits and a value of the one or more extra bits.