Home
> IP cores
> Arithmetic floating point > **Arithmetic floating point**

Floating point arithmetic | ACEN

Monday 14 October 2013, by

In this version the arithmetic IPs operate on 32 bits floating point numbers compliant with IEEE 754-1985. The error on the mantissa (with respect to the result calculated with an infinite precision) is less or equal to 0.5 bit. Normalized and de-normalized numbers defined by the standard are processed.

Arithmetic operation I | Data sheet | Evaluation kit for Lattice ECP3 | |
---|---|---|---|

Inversion of floating point numbers | |||

Multiplication of floating point numbers | |||

Division of floating point numbers | |||

Square root of floating point numbers | |||

Cube root of floating point numbers |