What Are the Limits to Quasi-CW in Non-Linear Optics?

[n0_3sc]

Could someone please explain what the limits to Quasi-CW is?
ie. If I have a train of square pulses, to what limit can I make the approximation that it can be treated as a Continuous Wave signal?

If that question was too vague here is my situation:
In Non Linear Optics you need lots of "watts" of power, now a 50W CW laser is impractical due to many reasons...so we use pulses...the shorter the pulse the higher the peak power.
However, most of Non Linear Optics theory relies on the fact that we are using CW lasers, but how short can you make a pulse until its no longer considered CW??

 

[JeffKoch]

I think it depends on the specific question you're asking about the effects of the laser. If you care about average power over a long time scale and don't care about the frequency with which the energy is delivered, then any frequency significantly higher than the inverse of this time scale is "Quasi-CW". What, specifically, is the question?

 

[n0_3sc]

The basic idea is I am doing a project involving a laser ring-cavity oscillator (more specifically a parametric oscillator), now because its made from only optical fibers the length of this oscillator is designed such that it is the length of a 1ns pulse at a repetition of 1Mhz. The average power is NOT important at all, we are concerned about peak powers.

So would it be correct by saying that it really doesn't matter on the repetition rate so long as this oscillators fiber length is adjusted to the right amount?