This talk explores upper and lower bounds on sampling rates for determining the homology of a maifold from noisy random samples.