Transmission line: 1/10th of the wavelength?

It says in a book that if the wire in a IC is greater than the 1/10th of the wavelength of the RF signal, then transmission line should be considered.

My question about the wire is:
Does it mean the total length of all the wires in a IC or just the longest wire in a IC?

Thank you!

This 1/10th rule of thumb applies to any wire (or trace) carrying RF inside of an IC.
Now there are IC's working at 60GHz for example, where any internal trace could be a design issue.
Some RF designers makes this rule even tougher: 1/16th

Thanks vfone! I am wondering about a case that if the total length of all RF signal wires on a RFIC is greater than 1/10th of the wavelength but the longest RF signal wire on the IC is less than 1/10th of the wavelength. How should I deal with this situation? Thanks!

By the way, the operation frequency I am using is 433MHz

You should think "signal path" rather than wire. If the signal path is short, we can use lumped model assumptions (Kirchhoff's law).
Otherwise, we have to consider transmission line effects (variation of current/voltage along the signal path etc.)

Thanks! So you mean the total length of all RF signal paths in the RF IC as long as RF signal goes though the path rather than the main signal path from the input to the output of the RF IC?

It is difficult to answer your question in such a general and absolute way.

These rules are based on "cascading 50 ohm building blocks with lines" thinking. In reality, it depends on how much effect one mismatched line will have on other blocks further down the signal path.

If we have a wire and then a series MIMCAP and then another wire, then we have to look at the total length of this path. For the signal, this is one transparent signal path. But if we have a line between two functional blocks (mixers, amplifiers, ...) then I would look at the length of this wire alone, if the functional blocks at the end of the wire "isolate" a possible mismatch from the rest of the circuit.

So there is some "gray zone" in this estimate, as you can see.

Also, this discussion about "short lines don't matter" is based on a medium impedance level in the signal path, such as cascading 50 ohm blocks.
If we look at very low or very high impedance levels, then even small capacitance or inductance values can be critical. Example: Adding a short line in the emitter/source (=extra inductance) is critical even though it is a short line, because here we have a low impedance level. Adding a wide line in a high impedance location will kill your performance because the little extra capacitance is critical at this impedance level. So the "short lines don't matter" is based on a medium impedance level in the signal path, such as cascading 50 ohm blocks.

Thank you. What do you mean by saying the functional blocks isolate a possible mismatch from the rest of the circuit? Could you give me more details? Or an example? Thank you very much.

The impedance of the signal path I have been trying to match up to is 50 ohms.