diff --git a/frontends/numpy-scipy/cometpy/MLIRGen/lowering.py b/frontends/numpy-scipy/cometpy/MLIRGen/lowering.py
index d7931c3c..84ac9a83 100644
--- a/frontends/numpy-scipy/cometpy/MLIRGen/lowering.py
+++ b/frontends/numpy-scipy/cometpy/MLIRGen/lowering.py
@@ -401,6 +401,7 @@ def translate_and_exec_llvm_with_jit(llvm_in,scf_lower_flags, func_name, inputs,
     # 2. Call mlir-translate to convert llvm to llvmir 
     # 3. Call clang to generate library
     # p = subprocess.run(to_llvm_command, input=llvm_in.encode('utf-8'), stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
+    print(llvm_in)
     p = subprocess.run(to_llvm_command +' 2>&1 |  '+ translate_mlir_command +' | ' + gcc_command , input=llvm_in.encode('utf-8'), stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
     if(p.returncode != 0):
         cleanup()
@@ -561,8 +562,8 @@ def lower_dialect_with_jit(ta_dialect_rep, target: str, out_dims, compile_with_f
             mlir_lower_flags += "--opt-fusion"
             compile_with_flags = compile_with_flags.replace("--opt-fusion","")
             compile_with_flags = compile_with_flags.replace("--opt-comp-workspace","")
-        if "-opt-matmul-tiling" not in compile_with_flags:
-            mlir_lower_flags += "   --convert-to-loops "
+        # if "-opt-matmul-tiling" not in compile_with_flags:
+        mlir_lower_flags += "   --convert-to-loops "
         mlir_lower_flags =" "+compile_with_flags + mlir_lower_flags
     else:
         mlir_lower_flags = "  --convert-ta-to-it --convert-to-loops "
diff --git a/lib/Dialect/TensorAlgebra/Transforms/LinalgTransforms.cpp b/lib/Dialect/TensorAlgebra/Transforms/LinalgTransforms.cpp
index 5e41ae84..b233d2d1 100644
--- a/lib/Dialect/TensorAlgebra/Transforms/LinalgTransforms.cpp
+++ b/lib/Dialect/TensorAlgebra/Transforms/LinalgTransforms.cpp
@@ -31,21 +31,27 @@
 
 #include "mlir/Dialect/Affine/IR/AffineOps.h"
 #include "mlir/Dialect/Affine/LoopUtils.h"
+#include "mlir/Dialect/Arith/IR/Arith.h"
 #include "mlir/Dialect/Linalg/IR/Linalg.h"
 #include "mlir/Dialect/Linalg/Utils/Utils.h"
 #include "mlir/Dialect/Linalg/Transforms/Transforms.h"
 #include "mlir/Dialect/Linalg/TransformOps/LinalgTransformOps.h"
+#include "mlir/Dialect/SCF/IR/SCF.h"
 #include "mlir/Dialect/SCF/Transforms/TileUsingInterface.h"
 #include "mlir/Dialect/Func/IR/FuncOps.h"
 #include "mlir/Dialect/LLVMIR/LLVMDialect.h"
 #include "mlir/Dialect/Tensor/IR/Tensor.h"
+#include "mlir/IR/BuiltinAttributes.h"
 #include "mlir/IR/BuiltinTypeInterfaces.h"
 #include "mlir/IR/BuiltinTypes.h"
+#include "mlir/IR/OpDefinition.h"
 #include "mlir/IR/PatternMatch.h"
 #include "mlir/IR/Value.h"
+#include "mlir/IR/ValueRange.h"
 #include "mlir/Pass/Pass.h"
 #include "mlir/Transforms/GreedyPatternRewriteDriver.h"
 #include "llvm/ADT/SmallVector.h"
+#include <optional>
 
 // suppress all warnings coming from inclusion of blis.h in source tree
 #ifdef __clang__
@@ -299,7 +305,7 @@ namespace
       get_level3_blocksizes(&mc, &kc, &nc, &mr, &nr, sizeof(double));
 
       addPatternForTiling(ctx, tilingPatterns, "__with_tiling__", "__L2__with_tiling__", {mc, nc, kc}, false, {1, 2, 0});
-      addPatternForTiling(ctx, tilingPatterns, "__L2__with_tiling__", "__micro_kernel__", {mr, nr, kc}, false, {1, 0, 2});
+      addPatternForTiling(ctx, tilingPatterns, "__L2__with_tiling__", "__micro_kernel__", {mr, nr, kc}, true, {1, 0, 2});
 
       if (failed(applyPatternsAndFoldGreedily(getOperation(),
                                               std::move(tilingPatterns))))
@@ -739,40 +745,32 @@ struct OptDenseTranspose : public ConversionPattern
       currentOrder.push_back(i);
     }
 
-    SmallVector<mlir::Value, 6> in_ivs;
-    SmallVector<mlir::Value, 6> out_ivs;
+    SmallVector<OpFoldResult, 6> in_ivs;
+    SmallVector<OpFoldResult, 6> out_ivs;
     in_ivs.resize(optimalOrder.size());
     out_ivs.resize(outputIndices[0].size());
-    mlir::Value carried_val = output;
-    SmallVector<AffineForOp, 6> loops;
-
+    OpFoldResult one = rewriter.createOrFold<ConstantIndexOp>(loc, 1);
+    SmallVector<OpFoldResult, 4> ubs;
     for (unsigned i = 0; i < optimalOrder.size(); i++)
     {
-      int64_t upperBound = inputType.getDimSize(optimalOrder[i]);
-      if (upperBound == ShapedType::kDynamic)
-      {
-        assert(false && "TODO: This dimension is a dynamic size");
-      }
-
-      /// create for loops
-      auto loop = rewriter.create<AffineForOp>(loc, 0, upperBound, 1, carried_val);
-      loops.push_back(loop);
-      rewriter.setInsertionPointToStart(loop.getBody());
-      in_ivs[optimalOrder[i]]  = loop.getInductionVar();
-      out_ivs[optimalOrder[outputIndices[0][i]]] = loop.getInductionVar();
-      carried_val = loop.getRegionIterArgs().front();
+      Value upperBound = rewriter.create<tensor::DimOp>(loc, input, optimalOrder[i]);
+      ubs.push_back(upperBound);
     }
-    
-    auto load_rhs = rewriter.create<tensor::ExtractOp>(loc, input, in_ivs);
-    auto store_lhs = rewriter.create<tensor::InsertOp>(loc, load_rhs, carried_val, out_ivs);
-    rewriter.create<AffineYieldOp>(loc, store_lhs.getResult());
-    for(int64_t i = loops.size()-2; i>=0 ; i--)
+
+    SmallVector<OpFoldResult, 4> ones(optimalOrder.size(), one);
+    auto forAll = rewriter.create<scf::ForallOp>(loc, ubs, output, std::nullopt);
+    rewriter.setInsertionPointToStart(forAll.getBody());
+    auto ivs = forAll.getLoopInductionVars();
+    for(size_t i = 0; i< forAll.getLoopInductionVars()->size(); i++)
     {
-      rewriter.setInsertionPointToEnd(loops[i].getBody());
-      rewriter.create<AffineYieldOp>(loc, loops[i+1].getResult(0));
+      in_ivs[optimalOrder[i]]  = forAll.getLoopInductionVars()->data()[i];
+      out_ivs[optimalOrder[outputIndices[0][i]]] = forAll.getLoopInductionVars()->data()[i];
     }
+    auto extracts = rewriter.create<tensor::ExtractSliceOp>(loc, input, in_ivs, ones, ones);
+    rewriter.setInsertionPointToEnd(forAll.getTerminator().getBody());
+    rewriter.create<tensor::ParallelInsertSliceOp>(loc, extracts, forAll.getRegionIterArgs().front(), out_ivs, ones, ones);
 
-    rewriter.replaceAllUsesWith(op->getResult(0), loops[0].getResult(0));
+    rewriter.replaceAllUsesWith(op->getResult(0), forAll->getResult(0));
     rewriter.eraseOp(op);
 
     //module.dump();
@@ -796,7 +794,7 @@ namespace
       comet_debug() << "OptDenseTransposePass : public PassWrapper<OptDenseTransposePass, FunctionPass>\n";
       func::FuncOp func = getOperation();
       ConversionTarget target(getContext());
-      target.addLegalDialect<TADialect, ArithDialect, AffineDialect, tensor::TensorDialect>();
+      target.addLegalDialect<TADialect, ArithDialect, scf::SCFDialect, AffineDialect, tensor::TensorDialect>();
       RewritePatternSet patterns(&getContext());
       patterns.insert<OptDenseTranspose>(&getContext(), tile_size, seperate_tiles);
 
diff --git a/test/integration/opts/ccsd_t1_21_ttgt_all_opts.ta b/test/integration/opts/ccsd_t1_21_ttgt_all_opts.ta
index 602631b0..2008e7ee 100644
--- a/test/integration/opts/ccsd_t1_21_ttgt_all_opts.ta
+++ b/test/integration/opts/ccsd_t1_21_ttgt_all_opts.ta
@@ -1,5 +1,5 @@
 # RUN: comet-opt  -opt-matmul-tiling -opt-matmul-mkernel -opt-dense-transpose --convert-tc-to-ttgt --convert-to-llvm %s &> ccsd_t1_21_ttgt_all.llvm
-# RUN: mlir-cpu-runner ccsd_t1_21_ttgt_all.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext | FileCheck %s
+# RUN: mlir-cpu-runner ccsd_t1_21_ttgt_all.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext,%mlir_utility_library_dir/libomp%shlibext | FileCheck %s
 
 def main() {
     #IndexLabel Declarations
diff --git a/test/integration/opts/ccsd_t1_21_ttgt_tiling.ta b/test/integration/opts/ccsd_t1_21_ttgt_tiling.ta
index 4d558506..f9b2166c 100644
--- a/test/integration/opts/ccsd_t1_21_ttgt_tiling.ta
+++ b/test/integration/opts/ccsd_t1_21_ttgt_tiling.ta
@@ -1,10 +1,10 @@
 # RUN: comet-opt --opt-matmul-tiling --convert-tc-to-ttgt --convert-to-llvm %s &> ccsd_t1_21_ttgt_tiling.llvm
-# RUN: mlir-cpu-runner ccsd_t1_21_ttgt_tiling.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext | FileCheck %s
+# RUN: mlir-cpu-runner ccsd_t1_21_ttgt_tiling.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext,%mlir_utility_library_dir/libomp%shlibext | FileCheck %s
 
 def main() {
     #IndexLabel Declarations
-    IndexLabel [i, c] = [2];
-    IndexLabel [m, n, a] = [4];
+    IndexLabel [i, c] = [16];
+    IndexLabel [m, n, a] = [32];
 
     Tensor<double> v([i, c, m, n], {Dense});
     Tensor<double> t2([m, n, c, a], {Dense});
@@ -21,4 +21,4 @@ def main() {
 
 # Print the result for verification.
 # CHECK: data = 
-# CHECK-NEXT: 250.24,250.24,250.24,250.24,250.24,250.24,250.24,250.24,
\ No newline at end of file
+# CHECK-NEXT: 128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,128123,
\ No newline at end of file
diff --git a/test/integration/opts/ccsd_t1_4_ttgt_bestperm.ta b/test/integration/opts/ccsd_t1_4_ttgt_bestperm.ta
index 64f6897c..5570422b 100644
--- a/test/integration/opts/ccsd_t1_4_ttgt_bestperm.ta
+++ b/test/integration/opts/ccsd_t1_4_ttgt_bestperm.ta
@@ -1,5 +1,5 @@
 # RUN: comet-opt --convert-tc-to-ttgt --convert-to-llvm %s &> ccsd_t1_4_ttgt_bestperm.llvm
-# RUN: mlir-cpu-runner ccsd_t1_4_ttgt_bestperm.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext | FileCheck %s
+# RUN: mlir-cpu-runner ccsd_t1_4_ttgt_bestperm.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext,%mlir_utility_library_dir/libomp%shlibext | FileCheck %s
 
 def main() {
     #IndexLabel Declarations
diff --git a/test/integration/opts/opt_dense_transpose.ta b/test/integration/opts/opt_dense_transpose.ta
index 6533fafe..01850fc3 100644
--- a/test/integration/opts/opt_dense_transpose.ta
+++ b/test/integration/opts/opt_dense_transpose.ta
@@ -1,5 +1,5 @@
 # RUN: comet-opt -opt-dense-transpose --convert-ta-to-it --convert-to-loops --convert-to-llvm %s &> opt_dense_transpose.llvm
-# RUN: mlir-cpu-runner opt_dense_transpose.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext | FileCheck %s
+# RUN: mlir-cpu-runner opt_dense_transpose.llvm -O3 -e main -entry-point-result=void -shared-libs=%comet_utility_library_dir/libcomet_runner_utils%shlibext,%mlir_utility_library_dir/libomp%shlibext | FileCheck %s
 
 #TODO(gkestor): read dense input from file