What are fixed-point DSP chips and floating-point DSP chips? What are the characteristics of each?

First of all, you must understand what fixed-point arithmetic and floating-point arithmetic are

You must know how the DSP chip handles decimals

This involves the determination of decimals. mark question. The precision of numerical values ??and the range of numerical values ??become a contradiction.

Bala, it’s a bit long, so I won’t go into it

You need to know: DSPs that support floating point operations are floating point DSP chips, and decimal operations can be performed freely on this chip. .

On the fixed-point DSP, decimal operations cannot be performed directly. The value needs to be calibrated first, and the decimal point cannot be moved.

Therefore, when fixed-point DSP completes decimal operations, the programming is slightly more complicated. The speed is definitely not as good as floating point movies.

Since floating point is better! Then it must be expensive too. 2812 (fixed point) 80 yuan 28335 (floating point) 100 yuan