Solution

ID: past-exam-of-the-mathematics-course-of-the-university-of-cambridge/2022/iii/paper-117/2/a/solution

The Mellin transform of is
Because the support is a compact subset of , the integral defines an entire function of . The Mellin inversion formula says that, for every real and every ,
Apply this with and initially . Absolute convergence of the Dirichlet series for the logarithmic derivative permits interchange of sum and integral, giving
Truncate at height , where is a sufficiently large fixed constant. The assumed bound on makes the discarded tails smaller than the required error. The classical Zero-free region of the Riemann zeta function and the standard bound there allow the truncated contour to move to
The only singularity crossed is the simple pole of at , whose residue is . Its contribution is
On the new contour, ; the zeta bounds, contour length, and exponential decay of absorb into a slight decrease of . Therefore
This is the smoothed prime number theorem from a zero-free region.

New to topics? Read the docs here!