I just purchased this MOSFET:
https://assets.nexperia.com/documents/data-sheet/PSMN017-30PL.pdf
and I am using it in conjunction with a microcontroller as a low side driver. The microcontroller outputs 488Hz pwm at 3.3V logic level. The input voltage to the mosfet is 24V and the output current is about ~1.2-1.3A. I am calculating temperature rise according the to the information in the data sheet choosing the least favorable numbers.
Power consumption: Rd-on * I^2 = 0.04536W
delta T = P*Rth = 2.72K
according to this the temperature of the MOSFET should only rise by about 2.72 degrees. I tried running the mosfet by connecting constant 3.3V to the gate so that I can ignore heating from switching. What I found was that the MOSFET heats up by a lot in a very short period of time. What am I missing?

 

What am I missing?

The Rds(on) at a Vgs of 3.3V.
It's specified at 4.5V:


Note the rapid increase in Rds(on) below a Vgs of 5V.
And that shows typical values. It could be worse for a worst-case device.


You need a MOSFET where the Rds(on) is rated at a Vgs of 3.3V or less.
Or you could add a transistor driver to increase Vgs to 5V or more (10V is optimum).

 

You can add a npn and a pnp bjt between uC and gate as the right top fig-07, and using a 10K to replace Rc2 and led, the last is to connects the Vc of Q2 to the gate of mosfet, the input is connects to the uC, and +Vcc connects to +5V.

And how is the load?

If the load is a coil and any kinds of inductance then you need to add a flywheel diode to protect the mosfet.

 

The Rds(on) at a Vgs of 3.3V.
It's specified at 4.5V:
).

Of various MOSFETs harvested from scrap PC motherboards - most will switch adequately with just under 2V on the gate.

The Vds is usually no more than 30V and sometimes as low as 20V, but the Id is usually at lest 80A.