Hi guys,
I have a project which needs me to create a VHDL code for a floating point multiplier using Quartus II.
As we know, the floating point can be represented in binary in 32 bits: 1 sign bit, 8 exponent bits and 23 mantissa bits.
The multiplier will need to accept two inputs: A and B.
The hints given for this prject is that I need to create an adder for the exponent and a multiplier for the mantissa.
I'm required to create my own multiplier. Creating a 23 bit multiplier for the mantissa is possible..but not easy. I'm confused whether should I use the shift method or the array method for the multiplication.
Besides that, do I really need to create a 23 bit multiplier just for the mantissa? Or is there any shorter way?
Any comments are appreciated. Thanks!
2 Replies  9146 Views  Last Post: 31 January 2011  07:59 PM
Replies To: VHDL Floating Point Multiplier
#2
Re: VHDL Floating Point Multiplier
Posted 09 January 2011  09:54 PM
You might have to look for other examples of how other programmers have done it.
Did you try writing your program with both the array method and the shift method? Use BBCode when posting. For example:
Did you try writing your program with both the array method and the shift method? Use BBCode when posting. For example:
Paste Your Code Hereand post both of your program examples so that we can compare them and look for a shorter way.
#3
Re: VHDL Floating Point Multiplier
Posted 31 January 2011  07:59 PM
I am doing a similar project this term, and right now designing a multiplier with shifting and adding. if you like, we can discuss and solve issues together.
cheers.
cheers.
deejay220989, on 09 January 2011  01:07 AM, said:
Hi guys,
I have a project which needs me to create a VHDL code for a floating point multiplier using Quartus II.
As we know, the floating point can be represented in binary in 32 bits: 1 sign bit, 8 exponent bits and 23 mantissa bits.
The multiplier will need to accept two inputs: A and B.
The hints given for this prject is that I need to create an adder for the exponent and a multiplier for the mantissa.
I'm required to create my own multiplier. Creating a 23 bit multiplier for the mantissa is possible..but not easy. I'm confused whether should I use the shift method or the array method for the multiplication.
Besides that, do I really need to create a 23 bit multiplier just for the mantissa? Or is there any shorter way?
Any comments are appreciated. Thanks!
I have a project which needs me to create a VHDL code for a floating point multiplier using Quartus II.
As we know, the floating point can be represented in binary in 32 bits: 1 sign bit, 8 exponent bits and 23 mantissa bits.
The multiplier will need to accept two inputs: A and B.
The hints given for this prject is that I need to create an adder for the exponent and a multiplier for the mantissa.
I'm required to create my own multiplier. Creating a 23 bit multiplier for the mantissa is possible..but not easy. I'm confused whether should I use the shift method or the array method for the multiplication.
Besides that, do I really need to create a 23 bit multiplier just for the mantissa? Or is there any shorter way?
Any comments are appreciated. Thanks!
Page 1 of 1
