diff --git a/WORKSPACE b/WORKSPACE
index b88b0d0b0..bbdbfed85 100644
--- a/WORKSPACE
+++ b/WORKSPACE
@@ -67,9 +67,9 @@ http_archive(
 
 http_archive(
     name = "qsim",
-    sha256 = "e2853379bde52d6277f9be4b80f54d32b3b27f7242a6c561cb34fb12d823b80e",
-    strip_prefix = "qsim-0.7.1-dev-20210126",
-    urls = ["https://github.com/quantumlib/qsim/archive/v0.7.1-dev+20210126.zip"],
+    sha256 = "5bc53d1d8c4689537d6f2cca20a67c8a1488f80a41cc412ca18d54afb6c21648",
+    strip_prefix = "qsim-0.8.0",
+    urls = ["https://github.com/quantumlib/qsim/archive/v0.8.0.zip"],
 )
 
 # Added for crosstool in tensorflow.
diff --git a/tensorflow_quantum/core/ops/tfq_calculate_unitary_op.cc b/tensorflow_quantum/core/ops/tfq_calculate_unitary_op.cc
index 6d444a829..0531d81db 100644
--- a/tensorflow_quantum/core/ops/tfq_calculate_unitary_op.cc
+++ b/tensorflow_quantum/core/ops/tfq_calculate_unitary_op.cc
@@ -106,19 +106,19 @@ class TfqCalculateUnitaryOp : public tensorflow::OpKernel {
 
     // Begin simulation.
     int largest_nq = 1;
-    Unitary u = UnitarySpace(largest_nq, tfq_for).CreateUnitary();
+    Unitary u = UnitarySpace(tfq_for).CreateUnitary(largest_nq);
 
     // Simulate programs one by one. Parallelizing over state vectors
     // we no longer parallelize over circuits. Each time we encounter a
     // a larger circuit we will grow the unitary as nescessary.
     for (int i = 0; i < fused_circuits.size(); i++) {
       int nq = num_qubits[i];
-      UCalculator sim = UCalculator(nq, tfq_for);
-      UnitarySpace us = UnitarySpace(nq, tfq_for);
+      UCalculator sim = UCalculator(tfq_for);
+      UnitarySpace us = UnitarySpace(tfq_for);
       if (nq > largest_nq) {
         // need to switch to larger unitaryspace.
         largest_nq = nq;
-        u = us.CreateUnitary();
+        u = us.CreateUnitary(nq);
       }
       us.SetIdentity(u);
       for (int j = 0; j < fused_circuits[i].size(); j++) {
@@ -136,7 +136,7 @@ class TfqCalculateUnitaryOp : public tensorflow::OpKernel {
           uint64_t k = l % (1 << max_num_qubits);
           if (k < crossover && j < crossover) {
             output_tensor(static_cast<ptrdiff_t>(i), static_cast<ptrdiff_t>(j),
-                          static_cast<ptrdiff_t>(k)) = us.GetEntry(u, j, k);
+                          static_cast<ptrdiff_t>(k)) = us.GetEntry(u, k, j);
           } else {
             output_tensor(static_cast<ptrdiff_t>(i), static_cast<ptrdiff_t>(j),
                           static_cast<ptrdiff_t>(k)) =