diff --git a/src/linalg/cholesky.rs b/src/linalg/cholesky.rs
index 312fd6cd..36e164ea 100644
--- a/src/linalg/cholesky.rs
+++ b/src/linalg/cholesky.rs
@@ -140,9 +140,9 @@ where
     }
 
     /// Computes the determinant of the decomposed matrix.
-    pub fn determinant(&self) -> N::SimdRealField {
+    pub fn determinant(&self) -> T::SimdRealField {
         let dim = self.chol.nrows();
-        let mut prod_diag = N::one();
+        let mut prod_diag = T::one();
         for i in 0..dim {
             prod_diag *= unsafe { *self.chol.get_unchecked((i, i)) };
         }
diff --git a/src/linalg/pow.rs b/src/linalg/pow.rs
index edcf55dd..33759a90 100644
--- a/src/linalg/pow.rs
+++ b/src/linalg/pow.rs
@@ -2,23 +2,27 @@
 
 use std::ops::DivAssign;
 
-use crate::{allocator::Allocator, DefaultAllocator, DimMin, MatrixN};
+use crate::{
+    allocator::Allocator,
+    storage::{Storage, StorageMut},
+    DefaultAllocator, DimMin, Matrix, OMatrix,
+};
 use num::PrimInt;
 use simba::scalar::ComplexField;
 
-impl<N: ComplexField, D> MatrixN<N, D>
+impl<T: ComplexField, D, S> Matrix<T, D, D, S>
 where
     D: DimMin<D, Output = D>,
-    DefaultAllocator: Allocator<N, D, D>,
-    DefaultAllocator: Allocator<N, D>,
+    S: StorageMut<T, D, D>,
+    DefaultAllocator: Allocator<T, D, D> + Allocator<T, D>,
 {
     /// Attempts to raise this matrix to an integral power `e` in-place. If this
     /// matrix is non-invertible and `e` is negative, it leaves this matrix
     /// untouched and returns `Err(())`. Otherwise, it returns `Ok(())` and
     /// overwrites this matrix with the result.
     #[must_use]
-    pub fn pow_mut<T: PrimInt + DivAssign>(&mut self, mut e: T) -> Result<(), ()> {
-        let zero = T::zero();
+    pub fn pow_mut<I: PrimInt + DivAssign>(&mut self, mut e: I) -> Result<(), ()> {
+        let zero = I::zero();
 
         // A matrix raised to the zeroth power is just the identity.
         if e == zero {
@@ -34,18 +38,19 @@ where
             }
         }
 
-        let one = T::one();
-        let two = T::from(2u8).unwrap();
+        let one = I::one();
+        let two = I::from(2u8).unwrap();
 
         // We use the buffer to hold the result of multiplier ^ 2, thus avoiding
         // extra allocations.
-        let mut multiplier = self.clone();
-        let mut buf = self.clone();
+        let mut multiplier = self.clone_owned();
+        let mut buf = self.clone_owned();
 
         // Exponentiation by squares.
         loop {
             if e % two == one {
-                *self *= &multiplier;
+                self.mul_to(&multiplier, &mut buf);
+                self.copy_from(&buf);
             }
 
             e /= two;
@@ -57,12 +62,20 @@ where
             }
         }
     }
+}
 
+impl<T: ComplexField, D, S: Storage<T, D, D>> Matrix<T, D, D, S>
+where
+    D: DimMin<D, Output = D>,
+    S: StorageMut<T, D, D>,
+    DefaultAllocator: Allocator<T, D, D> + Allocator<T, D>,
+{
     /// Attempts to raise this matrix to an integral power `e`. If this matrix
     /// is non-invertible and `e` is negative, it returns `None`. Otherwise, it
     /// returns the result as a new matrix. Uses exponentiation by squares.
-    pub fn pow<T: PrimInt + DivAssign>(&self, e: T) -> Option<Self> {
-        let mut clone = self.clone();
+    #[must_use]
+    pub fn pow<I: PrimInt + DivAssign>(&self, e: I) -> Option<OMatrix<T, D, D>> {
+        let mut clone = self.clone_owned();
 
         match clone.pow_mut(e) {
             Ok(()) => Some(clone),
diff --git a/tests/linalg/cholesky.rs b/tests/linalg/cholesky.rs
index bb578d49..6fd83912 100644
--- a/tests/linalg/cholesky.rs
+++ b/tests/linalg/cholesky.rs
@@ -92,7 +92,7 @@ macro_rules! gen_tests(
 
                 #[test]
                 fn cholesky_determinant_static(_n in PROPTEST_MATRIX_DIM) {
-                    let m = RandomSDP::new(U4, || random::<$scalar>().0).unwrap();
+                    let m = RandomSDP::new(Const::<4>, || random::<$scalar>().0).unwrap();
                     let lu_det = m.clone().lu().determinant();
                     assert_relative_eq!(lu_det.imaginary(), 0., epsilon = 1.0e-7);
                     let chol_det = m.cholesky().unwrap().determinant();