Guaranteed data compression
AssigneeImagination Technologies Limited
InventorSimon Fenney
Life of this patent
Estimating the term…
Abstract
Methods for converting an n-bit number into an m-bit number for situations where n>m and also for situations where n<m, where n and m are integers. The methods use truncation or bit replication followed by the calculation of an adjustment value which is applied to the replicated number.
Claims18
A method of mapping an input n-bit number to an output m-bit number, where n and m are integers and n>m, the method comprising:
- truncating, in a truncation hardware unit, the input n-bit number from n-bits to m-bits to form an intermediate m-bit number;
- in a look-up logic hardware unit:
- (i) comparing pre-determined subsets of the bits of the input n-bit number with pre-determined values in fixed-function circuitry, and (ii) setting an adjustment value in dependence on the results of the comparisons;
- and
- adding, in an increment/decrement hardware unit, the adjustment value to the intermediate m-bit number to generate the output m-bit number.
A method of mapping an input n-bit number to an output m-bit number, where n and m are integers and n<m, the method comprising:
- appending, in a replication hardware unit, (m−n) most significant bits of the input n-bit number to the input n-bit number to form an intermediate m-bit number;
- in a look-up logic hardware unit:
- (i) comparing pre-determined subsets of the bits of the input n-bit number with pre-determined values in fixed-function circuitry, and (ii) setting an adjustment value in dependence on the results of the comparisons;
- and
- adding, in an increment/decrement hardware unit, the adjustment value to the intermediate m-bit number to generate the output m-bit number.
Hardware logic arranged to map an input n-bit number to an output m-bit number, where n and m are integers and n>m, the hardware logic comprising:
- a truncation hardware unit arranged to truncate the input n-bit number from n-bits to m-bits to form an intermediate m-bit number;
- a look-up logic hardware unit arranged to (i) compare pre-determined subsets of the bits of the input n-bit number with pre-determined values in fixed-function circuitry, and (ii) set an adjustment value in dependence on the results of the comparisons;
- and
- an increment/decrement hardware unit arranged to add the adjustment value to the intermediate m-bit number to generate the output m-bit number.
Dependencies are inferred from the claim text.