Real Random Number Generator

[DeaD_EyE]

You miss the point. Digital computers do it very well and fast if done correctly.

Knowing the initial-state, kills pseudo random number generators.

import random

random.seed(1337) # <- initial state
# is used for unit tests

print([random.randint(0, 10) for _ in range(10)])

Output:
[9, 8, 5, 9, 9, 2, 5, 6, 10, 5]

Real entropy comes from nature. Everything else, is not real entropy.
Another example of an entropy source is an ADC with noise.

But be aware of side channel attacks.
If you can fake or listen to an entropy source, it's possible to calculate all values from PSRNG.

Desktop PCs do have many entropy sources to improve the situation.
Embedded systems do not have as many entropy sources as a Desktop PC.

This is a problem, if cryptographic algorithms are used, which require a high entropy.