MVUE for $\sigma^2$ — known mean removes the Bessel correction

Let $X_1,X_2,\dots,X_n$ be independently and normally distributed with mean 0 and variance $\sigma^2$. What is the minimum variance unbiased estimator of $\sigma^2$?