How does Arduino handle floating-point operations?

Yes Arduino can do floating-point operations, but there’s a few things to keep in mind.

1. On boards like the Uno or Nano (which use the ATmega328P), it supports float and double, but the funny thing is—they’re actually the same thing. Both are 32-bit IEEE 754 floating point numbers. So don’t expect extra precision with double, it's just a float behind the scenes.

2. It can handle basic operations like addition, subtraction, multiplication, and division just fine. But it’s not super fast at it, since the 8-bit microcontrollers don’t have a floating point unit (FPU). That means it does all floating-point math in software, which can slow things down if you’re doing a lot of calculations in your loop.

3. Also, things like sin(), cos(), sqrt() and pow() work, but again, they’re kinda heavy on processing time. So if you’re working with sensors and need to process stuff quickly, it’s sometimes better to stick with integers where possible, or scale up the values and work in "fixed point" math if you can.

4. One more thing—printing floats with Serial.print() only shows two decimal places by default. You can control that though:

float pi = 3.14159;
Serial.print(pi);         // prints 3.14
Serial.println(pi, 4);    // prints 3.1416

So yeah, Arduino can handle floats, but it’s not optimized for heavy-duty number crunching. If you're doing more advanced math or need higher precision, better to move to something like a Teensy or a 32-bit board like the Arduino Due or even ESP32. But for basic stuff, it’s totally usable.