Suppose a floating point format has an n-bit frac field and a k bit exp field. What is the smallest integer that can be represented exactly? Assume that k is large enough not to provide any limitations on the problem. What is the integer for IEEE-754 single precision?