how to deal with the 2000mA peak current demand of a cellular shield

I am running the Sparkfun 5100B cellular shield on an Uno.

After some initial problems, it's now all running fine, and I've ssuccessfully sent and received SMS messages.

I'm now trying to finalize the power supply to my set-up, and I notice from the data sheets that the cellular shield idles at about 7mA, but peaks at an enormous 2000mA when transmitting. My question is how "serious" this limit is. I note that:

(a) the UNO/Cellular shield I used for testing is powered by USB from a PC. The nominal USB hub limit is 500mA, nowhere near the 2000mA. I would imagine that there are other nominal limits on the various components of the power 'train' within the Arduino itself which are also being exceded.

(b) the shield is designed to mate with the Uno, and sold without any loud warnings about definitely needing to beef up the power supply to 2000mA.

(c) my Uno is not fried - yet!

What is going on here?

  • Will the PC give 2000mA for a short burst without melting/complaining?
  • Will the Arduino pass on 2000mA for a short burst without melting/complaining
  • Is the 2000mA an exaggeration?
  • Is there some kind of capacitance within the system that facilitates short power surges?

Behind these questions, I'm obviously afraid that I've just been lucky so far and that I am courting disaster.

Have you used such a shield? Did you beef up your power supply?

Please enlighten me.

Thanks

Kenny

Will the PC give 2000mA for a short burst without melting/complaining?

No. 500mA is a hard limit.

Will the Arduino pass on 2000mA for a short burst without melting/complaining

No.

Is the 2000mA an exaggeration?

A worst-case scenario, most likely. And probably high, to cover their posteriors.

Is there some kind of capacitance within the system that facilitates short power surges?

No. The amount of current available is limited. If a device needs more than that, it doesn't get it.