diff --git a/barretenberg/cpp/scripts/benchmark_example_ivc_flow_remote.sh b/barretenberg/cpp/scripts/benchmark_example_ivc_flow_remote.sh
index 514240c14af3..174f42f13640 100755
--- a/barretenberg/cpp/scripts/benchmark_example_ivc_flow_remote.sh
+++ b/barretenberg/cpp/scripts/benchmark_example_ivc_flow_remote.sh
@@ -7,7 +7,7 @@ TARGET=${1:-"bb"}
 #FLOW=${2:-"ecdsar1+transfer_1_recursions+sponsored_fpc"}
 #FLOW=${2:-"ecdsar1+transfer_1_recursions+sponsored_fpc"}
 FLOW=${2:-"schnorr+deploy_tokenContract_with_registration+sponsored_fpc"}
-BUILD_DIR="build-no-avm"
+BUILD_DIR="build"
 
 # Move above script dir.
 cd $(dirname $0)/..
diff --git a/barretenberg/cpp/src/barretenberg/honk/composer/permutation_lib.hpp b/barretenberg/cpp/src/barretenberg/honk/composer/permutation_lib.hpp
index 619843b7753a..4ffb8ff61850 100644
--- a/barretenberg/cpp/src/barretenberg/honk/composer/permutation_lib.hpp
+++ b/barretenberg/cpp/src/barretenberg/honk/composer/permutation_lib.hpp
@@ -125,6 +125,8 @@ PermutationMapping<Flavor::NUM_WIRES> compute_permutation_mapping(
     const std::vector<CyclicPermutation>& wire_copy_cycles)
 {
 
+    BB_BENCH_NAME("compute_permutation_mapping");
+
     // Initialize the table of permutations so that every element points to itself
     PermutationMapping<Flavor::NUM_WIRES> mapping(dyadic_size);
 
@@ -132,43 +134,45 @@ PermutationMapping<Flavor::NUM_WIRES> compute_permutation_mapping(
     std::span<const uint32_t> real_variable_tags = circuit_constructor.real_variable_tags;
 
     // Go through each cycle
-    for (size_t cycle_idx = 0; cycle_idx < wire_copy_cycles.size(); ++cycle_idx) {
-        // We go through the cycle and fill-out/modify `mapping`. Following the generalized permutation algorithm, we
-        // take separate care of first/last node handling.
-        const CyclicPermutation& cycle = wire_copy_cycles[cycle_idx];
-        const auto cycle_size = cycle.size();
-        if (cycle_size == 0) {
-            continue;
-        }
+    parallel_for([&](const ThreadChunk& chunk) {
+        for (size_t cycle_idx : chunk.range(wire_copy_cycles.size())) {
+            // We go through the cycle and fill-out/modify `mapping`. Following the generalized permutation
+            // algorithm, we take separate care of first/last node handling.
+            const CyclicPermutation& cycle = wire_copy_cycles[cycle_idx];
+            const auto cycle_size = cycle.size();
+            if (cycle_size == 0) {
+                continue;
+            }
 
-        const cycle_node& first_node = cycle[0];
-        const cycle_node& last_node = cycle[cycle_size - 1];
+            const cycle_node& first_node = cycle[0];
+            const cycle_node& last_node = cycle[cycle_size - 1];
 
-        const auto first_row = static_cast<ptrdiff_t>(first_node.gate_idx);
-        const auto first_col = first_node.wire_idx;
-        const auto last_row = static_cast<ptrdiff_t>(last_node.gate_idx);
-        const auto last_col = last_node.wire_idx;
+            const auto first_row = static_cast<ptrdiff_t>(first_node.gate_idx);
+            const auto first_col = first_node.wire_idx;
+            const auto last_row = static_cast<ptrdiff_t>(last_node.gate_idx);
+            const auto last_col = last_node.wire_idx;
 
-        // First node: id gets tagged with the cycle's variable tag
-        mapping.ids[first_col].is_tag[first_row] = true;
-        mapping.ids[first_col].row_idx[first_row] = real_variable_tags[cycle_idx];
+            // First node: id gets tagged with the cycle's variable tag
+            mapping.ids[first_col].is_tag[first_row] = true;
+            mapping.ids[first_col].row_idx[first_row] = real_variable_tags[cycle_idx];
 
-        // Last node: sigma gets tagged and points to tau(tag) instead of wrapping to first node
-        mapping.sigmas[last_col].is_tag[last_row] = true;
-        mapping.sigmas[last_col].row_idx[last_row] = circuit_constructor.tau().at(real_variable_tags[cycle_idx]);
+            // Last node: sigma gets tagged and points to tau(tag) instead of wrapping to first node
+            mapping.sigmas[last_col].is_tag[last_row] = true;
+            mapping.sigmas[last_col].row_idx[last_row] = circuit_constructor.tau().at(real_variable_tags[cycle_idx]);
 
-        // All nodes except the last: sigma points to the next node in the cycle
-        for (size_t node_idx = 0; node_idx + 1 < cycle_size; ++node_idx) {
-            const cycle_node& current_node = cycle[node_idx];
-            const cycle_node& next_node = cycle[node_idx + 1];
+            // All nodes except the last: sigma points to the next node in the cycle
+            for (size_t node_idx = 0; node_idx + 1 < cycle_size; ++node_idx) {
+                const cycle_node& current_node = cycle[node_idx];
+                const cycle_node& next_node = cycle[node_idx + 1];
 
-            const auto current_row = static_cast<ptrdiff_t>(current_node.gate_idx);
-            const auto current_col = current_node.wire_idx;
-            // Point current node to next node.
-            mapping.sigmas[current_col].row_idx[current_row] = next_node.gate_idx;
-            mapping.sigmas[current_col].col_idx[current_row] = static_cast<uint8_t>(next_node.wire_idx);
+                const auto current_row = static_cast<ptrdiff_t>(current_node.gate_idx);
+                const auto current_col = current_node.wire_idx;
+                // Point current node to next node.
+                mapping.sigmas[current_col].row_idx[current_row] = next_node.gate_idx;
+                mapping.sigmas[current_col].col_idx[current_row] = static_cast<uint8_t>(next_node.wire_idx);
+            }
         }
-    }
+    });
 
     // Add information about public inputs so that the cycles can be altered later; See the construction of the
     // permutation polynomials for details. This _only_ effects sigma_0, the 0th sigma polynomial, as the structure of
diff --git a/barretenberg/cpp/src/barretenberg/trace_to_polynomials/trace_to_polynomials.cpp b/barretenberg/cpp/src/barretenberg/trace_to_polynomials/trace_to_polynomials.cpp
index 4e062686edd2..8a82d30c2502 100644
--- a/barretenberg/cpp/src/barretenberg/trace_to_polynomials/trace_to_polynomials.cpp
+++ b/barretenberg/cpp/src/barretenberg/trace_to_polynomials/trace_to_polynomials.cpp
@@ -79,13 +79,15 @@ std::vector<CyclicPermutation> TraceToPolynomials<Flavor>::populate_wires_and_se
         RefVector<Selector<FF>> block_selectors = block.get_selectors();
         // Insert the selector values for this block into the selector polynomials at the correct offset
         // TODO(https://github.com/AztecProtocol/barretenberg/issues/398): implicit arithmetization/flavor consistency
-        for (size_t selector_idx = 0; selector_idx < block_selectors.size(); selector_idx++) {
-            auto& selector = block_selectors[selector_idx];
-            for (size_t row_idx = 0; row_idx < block_size; ++row_idx) {
-                size_t trace_row_idx = row_idx + offset;
-                selectors[selector_idx].set_if_valid_index(trace_row_idx, selector[row_idx]);
+        parallel_for([&](const ThreadChunk& chunk) {
+            for (size_t selector_idx = 0; selector_idx < block_selectors.size(); selector_idx++) {
+                auto& selector = block_selectors[selector_idx];
+                for (size_t row_idx : chunk.range(block_size)) {
+                    size_t trace_row_idx = row_idx + offset;
+                    selectors[selector_idx].set_if_valid_index(trace_row_idx, selector[row_idx]);
+                }
             }
-        }
+        });
     }
 
     return copy_cycles;