Scaling accumulator mult (signed value) in Distributed Arith

M

MariuszK

Guest
Hello,
I am going write distributed arithmetic unit for signed values.
Currently I have problem with scaling accumulator multiplier for
signed values (negative positive).

I use sign extending to input and table values to perform scaling
accumulation.
That's works but ... this is inefficient all additions must be double-
precision and at least twice as many partial products are needed.

Is it possibility to implement Scaling Accumulator Multipliers (for
signed values) more efficient? (without sign extending)

Regards,
Mariusz
 

Welcome to EDABoard.com

Sponsor

Back
Top