I am simulating an nxn Wallace Tree multiplier(without Booth's recording) that groups partial products in threes and uses 3:2 compressors for reducing them simultaneously in each step.
In the simulation I also compute the total number of adders for any nxn. Here is the result of one such test that I ran -
3X3, No. of Adders = 3 4X4, No. of Adders = 8 5X5, No. of Adders = 15 6X6, No. of Adders = 26 7X7, No. of Adders = 39 8X8, No. of Adders = 53 9X9, No. of Adders = 71 10X10, No. of Adders = 92 11X11, No. of Adders = 112 12X12, No. of Adders = 136 13X13, No. of Adders = 161 14X14, No. of Adders = 193 15X15, No. of Adders = 222 16X16, No. of Adders = 253 17X17, No. of Adders = 287 18X18, No. of Adders = 325 19X19, No. of Adders = 364 20X20, No. of Adders = 413 21X21, No. of Adders = 455 22X22, No. of Adders = 497 23X23, No. of Adders = 542 24X24, No. of Adders = 590 25X25, No. of Adders = 641 26X26, No. of Adders = 695 27X27, No. of Adders = 752 28X28, No. of Adders = 809 29X29, No. of Adders = 884 30X30, No. of Adders = 944 31X31, No. of Adders = 1004 32X32, No. of Adders = 1066Can anybody please make me understand as to why the number of adders increases so rapidly that it becomes more that n*n for lets say n=32 ??
Regards, Prasanna
--------------------------------------- This message was sent using the comp.arch.embedded web interface on
formatting link