Programming issues with Zhejiang University JX—300X
Thread Content
I used the Zhejiang University JX—300X to write a program for flow accumulation; the ‘INT_TO_SFLOAT’ module was employed to convert 3600 into a half-precision floating-point value. The input value was set to 3600, and the output was designated as TEMP1. However, an error related to the output pin was reported during compilation. Could someone explain this?
For the half-precision floating-point numbers used by Zhejiang University’s Zhongkong system, the range is from negative several to positive several; I can’t remember the exact values, but it’s definitely no more than 5. Converting values on such a large scale will cause an overflow
Using a mage for this is incorrect; one wants to convert 3600 into a half-precision floating-point number. First, the variable TEMP1 that you define must be a half-precision floating-point number: its range is 0 to 3600. Secondly, you need to convert INT to float first, then divide by the range of 3600, and finally convert the result to a half-precision floating-point number