Browse Prior Art Database

Fast and Modular 2048-Bit to 11-Bit Priority Encoder

IP.com Disclosure Number: IPCOM000110828D
Original Publication Date: 1994-Jan-01
Included in the Prior Art Database: 2005-Mar-26
Document File: 6 page(s) / 109K

Publishing Venue

IBM

Related People

Patel, M: AUTHOR

Abstract

Disclosed is a 2048-bit to 11-bit priority encoder that is fast and does not require high speed clocks required by sequential solutions.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Fast and Modular 2048-Bit to 11-Bit Priority Encoder

      Disclosed is a 2048-bit to 11-bit priority encoder that is fast
and does not require high speed clocks required by sequential
solutions.

      A priority encoder generates the location of the bit in the
input that has a one in the lowest address.  For example, DIN
(2047:0) is a vector of 2048 bits (bit 0 is least significant and bit
2047 is most significant).  The 11-bit priority encoded value, PE
(10:0) (bit 0 is least significant and bit 10 is most significant),
for DIN (2047:0) = XXX...XXX10000000 is PE (10:0) = 00000000111.

      There are sequential methods which require very fast clocks to
iteratively compute one bit at a time for computing the 11-bit
priority encoded number from 2048-bit value.  The problem is solved
for groups of 16-bits first and then the results of the first stage
are used to compute which group is to be selected for the second
stage and so on.  Fig. 1 shows a block diagram of a 2048-bit to
11-bit priority encoder.  It is made up of eight 256-bit to 8-bit
priority encoders (PE8), Fig. 2, that compute individual values
simultaneously and then one of the eight 8-bit result is selected via
the PSEL3 to select the lowest address PE8 that had a bit on.  An
8-bit to 3-bit priority encoder (PE3), see Fig. 3, is used to compute
the value of the lowest address PE8 that had a bit on.  The PE3
outputs form the high order 3 bits and the PSEL3 outputs form the
lower order 8 bits of the 11-bit prioritized value of the original
2048-bit value.

      PE3 and PE4 are well known priority encode functions so
truth-table description is provided for PE3 in Fig. 3.  PE4 is
similar to PE3 but for 4 bits.  PCNTL3, Fig. 5, describes the
truth-table for the priori...