Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Multiple Precision Pipelined Multiplier

IP.com Disclosure Number: IPCOM000093385D
Original Publication Date: 1967-Sep-01
Included in the Prior Art Database: 2005-Mar-06
Document File: 3 page(s) / 50K

Publishing Venue

IBM

Related People

Senzig, DN: AUTHOR

Abstract

The multiplier generates both single precision and double precision products without iteration. The transition from single precision to double precision is accomplished by recoding the multiplier in a different radix.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 56% of the total text.

Page 1 of 3

Multiple Precision Pipelined Multiplier

The multiplier generates both single precision and double precision products without iteration. The transition from single precision to double precision is accomplished by recoding the multiplier in a different radix.

The word length is either seventeen or thirty-four binary bits. Carry-assimilate adders 10...12 and shift circuitry, not shown, are used in double precision to generate seventeen different multiples of the multiplicand ranging from a multiple of minus eight times to plus eight times.

The multiplication tree includes carry-save adders 20...26 and a carry- assimilate adder 27. The inputs to the multiplication tree are designated A...I.

When operating in a single precision mode, the seventeen binary bits in the multiplier are recoded into nine signals having a radix of four. Each two adjacent bits in the multiplier are recoded into digits of value -2, -1, 0, +1, +2. The nine recoded multiplier digits are used to gate multiples of the multiplicand into inputs A...I of the multiplier tree. The bits from the multiplier which control the various inputs to the multiplication tree are tabulated below. Bits from 1 2&3 4&5 6&7 8&9 10&11 12&13 14&15 16&17

multiplier

Inputs To A B C D E F G H I

Tree Which

Are

Controlled.

In the case of a double precision multiplication, the multiplier has thirty-four binary bits. The same hardware is used with the exception that in the case of a double precision operation, the thirty-four bits in the multiplier are recoded i...