Give an estimate of the maximum number of arithmetic operations required in a fast exponentiation when the exponent has 100 decimal digits.
talked to prof about this review question, didn't really understand the reply, it was along the lines of, 2100 = 1000 digits = 103 x 10 = 330 binary digits x 2 = 660 operations , just hoping someone could explain the steps to get to 660 operations
The reasoning is probably that a number with $100$ decimal digits will have about $330$ bits -- and using exponentiation by squaring, you will need either one or two modular multiplications per bit position in the exponent, for a worst-case of about $660$ modular multiplications.
This does not count how much work it is to do one modular multiplication (which generally involves one bignum multiplication and a division with remainder, unless you use something like Montgomery reduction), nor does it take into account that squarings of large numbers can be done somewhat cheaper than general multiplications.