Hi All,
I have a 2 questions
1. What are the most critical maths you need to know when going for a DSP + FPGA interview ? , e.g multiplying 2 N bit numbers will take up 2* N bits.
2. How many bits are needed for a MAC unit ( say you have two 16 bits inputs to a multiplier which goes in the a accumulator of 40 bits) . I dont understand how the extra 8 bits ( 40 bits - 32 bits ) allow us to do add up to 256 32 bit numbers.
A good example is provided here **broken link removed** ,but i dont really understand how the generalized formula is contructed for a N bit MAC , adding M times.
Any help is appreciated