diff --git a/attic/src/integrator/cuda/cuda_eval_denvars.cu b/attic/src/integrator/cuda/cuda_eval_denvars.cu
index 0f1cd2f0c..c5f49ffa6 100644
--- a/attic/src/integrator/cuda/cuda_eval_denvars.cu
+++ b/attic/src/integrator/cuda/cuda_eval_denvars.cu
@@ -31,7 +31,7 @@ __global__ void eval_uvars_lda_kernel( size_t           ntasks,
   const int tid_x = blockIdx.x * blockDim.x + threadIdx.x;
   const int tid_y = blockIdx.y * blockDim.y + threadIdx.y;
 
-  register double den_reg = 0.;
+  double den_reg = 0.;
 
   if( tid_x < nbf and tid_y < npts ) {
 
@@ -115,10 +115,10 @@ __global__ void eval_uvars_gga_kernel( size_t           ntasks,
 
       for (int sm_y = threadIdx.y; sm_y < GGA_KERNEL_SM_BLOCK_Y; sm_y += blockDim.y) {
         const int tid_y = bid_y + sm_y;
-        register double den_reg = den_shared[0][sm_y][threadIdx.x];
-        register double dx_reg  = den_shared[1][sm_y][threadIdx.x];
-        register double dy_reg  = den_shared[2][sm_y][threadIdx.x];
-        register double dz_reg  = den_shared[3][sm_y][threadIdx.x];
+        double den_reg = den_shared[0][sm_y][threadIdx.x];
+        double dx_reg  = den_shared[1][sm_y][threadIdx.x];
+        double dy_reg  = den_shared[2][sm_y][threadIdx.x];
+        double dz_reg  = den_shared[3][sm_y][threadIdx.x];
 
         // Warp blocks are stored col major
         den_reg = 2 * warpReduceSum( den_reg );
diff --git a/attic/src/integrator/cuda/cuda_weights.cu b/attic/src/integrator/cuda/cuda_weights.cu
index 3013324b0..94ab86e9c 100644
--- a/attic/src/integrator/cuda/cuda_weights.cu
+++ b/attic/src/integrator/cuda/cuda_weights.cu
@@ -30,7 +30,7 @@ __global__ void compute_point_center_dist(
 ) {
 
   __shared__ double3 point_buffer[warp_size];
-  register double3 coord_reg;
+  double3 coord_reg;
 
   const int natoms_block = (natoms + warp_size-1) / warp_size;
   const int coords_block = (npts + warp_size-1) / warp_size;
diff --git a/attic/src/integrator/cuda/gauxc-cuda_integrator.cmake b/attic/src/integrator/cuda/gauxc-cuda_integrator.cmake
index 864739b9b..20d957ce7 100644
--- a/attic/src/integrator/cuda/gauxc-cuda_integrator.cmake
+++ b/attic/src/integrator/cuda/gauxc-cuda_integrator.cmake
@@ -32,7 +32,7 @@ target_sources( gauxc PRIVATE cuda/collocation_device.cu
 			      cuda/cuda_device_properties.cxx
 )
 
-target_compile_features( gauxc PRIVATE cuda_std_14 )
+target_compile_features( gauxc PRIVATE cuda_std_17 )
 #target_include_directories( gauxc
 #  PRIVATE
 #    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/src/integrator/cuda>
diff --git a/attic/src/new_integrator/device/cuda/cuda_eval_denvars.cu b/attic/src/new_integrator/device/cuda/cuda_eval_denvars.cu
index 813634370..54c17bb25 100644
--- a/attic/src/new_integrator/device/cuda/cuda_eval_denvars.cu
+++ b/attic/src/new_integrator/device/cuda/cuda_eval_denvars.cu
@@ -31,7 +31,7 @@ __global__ void eval_uvars_lda_kernel( size_t           ntasks,
   const int tid_x = blockIdx.x * blockDim.x + threadIdx.x;
   const int tid_y = blockIdx.y * blockDim.y + threadIdx.y;
 
-  register double den_reg = 0.;
+  double den_reg = 0.;
 
   if( tid_x < nbf and tid_y < npts ) {
 
@@ -115,10 +115,10 @@ __global__ void eval_uvars_gga_kernel( size_t           ntasks,
 
       for (int sm_y = threadIdx.y; sm_y < GGA_KERNEL_SM_BLOCK_Y; sm_y += blockDim.y) {
         const int tid_y = bid_y + sm_y;
-        register double den_reg = den_shared[0][sm_y][threadIdx.x];
-        register double dx_reg  = den_shared[1][sm_y][threadIdx.x];
-        register double dy_reg  = den_shared[2][sm_y][threadIdx.x];
-        register double dz_reg  = den_shared[3][sm_y][threadIdx.x];
+        double den_reg = den_shared[0][sm_y][threadIdx.x];
+        double dx_reg  = den_shared[1][sm_y][threadIdx.x];
+        double dy_reg  = den_shared[2][sm_y][threadIdx.x];
+        double dz_reg  = den_shared[3][sm_y][threadIdx.x];
 
         // Warp blocks are stored col major
         den_reg = 2 * warpReduceSum( den_reg );
diff --git a/attic/src/new_integrator/device/cuda/cuda_weights.cu b/attic/src/new_integrator/device/cuda/cuda_weights.cu
index f8da49bfc..98bae8268 100644
--- a/attic/src/new_integrator/device/cuda/cuda_weights.cu
+++ b/attic/src/new_integrator/device/cuda/cuda_weights.cu
@@ -30,7 +30,7 @@ __global__ void compute_point_center_dist(
 ) {
 
   __shared__ double3 point_buffer[warp_size];
-  register double3 coord_reg;
+  double3 coord_reg;
 
   const int natoms_block = (natoms + warp_size-1) / warp_size;
   const int coords_block = (npts + warp_size-1) / warp_size;
diff --git a/attic/src/new_integrator/device/cuda/gauxc-cuda.cmake b/attic/src/new_integrator/device/cuda/gauxc-cuda.cmake
index e95000a37..e2bb0e328 100644
--- a/attic/src/new_integrator/device/cuda/gauxc-cuda.cmake
+++ b/attic/src/new_integrator/device/cuda/gauxc-cuda.cmake
@@ -22,7 +22,7 @@ target_sources( gauxc PRIVATE
 
 )
 
-target_compile_features( gauxc PRIVATE cuda_std_14 )
+target_compile_features( gauxc PRIVATE cuda_std_17 )
 target_compile_options( gauxc
   PRIVATE
     $<$<COMPILE_LANGUAGE:CUDA>: -Xcudafe --diag_suppress=partial_override -Xptxas -v > 
diff --git a/src/runtime_environment/device/cuda/CMakeLists.txt b/src/runtime_environment/device/cuda/CMakeLists.txt
index 50ea945c4..18c3039b5 100644
--- a/src/runtime_environment/device/cuda/CMakeLists.txt
+++ b/src/runtime_environment/device/cuda/CMakeLists.txt
@@ -13,7 +13,7 @@ if( NOT TARGET CUDA::cublas )
   find_package( CUDAToolkit REQUIRED )
 endif()
 
-target_compile_features( gauxc PRIVATE cuda_std_14 )
+target_compile_features( gauxc PRIVATE cuda_std_17 )
 target_sources( gauxc PRIVATE cuda_backend.cxx )
 
 if(NOT GAUXC_LINK_CUDA_STATIC)
diff --git a/src/xc_integrator/local_work_driver/device/cuda/kernels/grid_to_center.cu b/src/xc_integrator/local_work_driver/device/cuda/kernels/grid_to_center.cu
index 03600010c..f90e2056d 100644
--- a/src/xc_integrator/local_work_driver/device/cuda/kernels/grid_to_center.cu
+++ b/src/xc_integrator/local_work_driver/device/cuda/kernels/grid_to_center.cu
@@ -28,7 +28,7 @@ __global__ void compute_grid_to_center_dist(
 ) {
 
   __shared__ double3 point_buffer[cuda::warp_size];
-  register double3 coord_reg;
+  double3 coord_reg;
 
   const int natoms_block = (natoms + cuda::warp_size-1) / cuda::warp_size;
   const int coords_block = (npts + cuda::warp_size-1) / cuda::warp_size;
diff --git a/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_gga.hpp b/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_gga.hpp
index 9b466e24a..be703841c 100644
--- a/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_gga.hpp
+++ b/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_gga.hpp
@@ -131,10 +131,10 @@ __global__ void eval_vvar_gga_kern( size_t        ntasks,
 
       for (int sm_y = threadIdx.y; sm_y < VVAR_KERNEL_SM_BLOCK; sm_y += blockDim.y) {
         const int tid_y = bid_y + sm_y;
-        register double den_reg = den_shared[0][sm_y][threadIdx.x];
-        register double dx_reg  = den_shared[1][sm_y][threadIdx.x];
-        register double dy_reg  = den_shared[2][sm_y][threadIdx.x];
-        register double dz_reg  = den_shared[3][sm_y][threadIdx.x];
+        double den_reg = den_shared[0][sm_y][threadIdx.x];
+        double dx_reg  = den_shared[1][sm_y][threadIdx.x];
+        double dy_reg  = den_shared[2][sm_y][threadIdx.x];
+        double dz_reg  = den_shared[3][sm_y][threadIdx.x];
 
         // Warp blocks are stored col major
         den_reg =     cuda::warp_reduce_sum<warp_size>( den_reg );
diff --git a/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_lda.hpp b/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_lda.hpp
index 54dc5043c..344acb6c6 100644
--- a/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_lda.hpp
+++ b/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_lda.hpp
@@ -49,7 +49,7 @@ __global__ void eval_vvar_lda_kern( size_t        ntasks,
   const int tid_x = blockIdx.x * blockDim.x + threadIdx.x;
   const int tid_y = blockIdx.y * blockDim.y + threadIdx.y;
 
-  register double den_reg = 0.;
+  double den_reg = 0.;
 
   if( tid_x < nbf and tid_y < npts ) {
 
diff --git a/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_mgga.hpp b/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_mgga.hpp
index 82b5207e1..427a5923a 100644
--- a/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_mgga.hpp
+++ b/src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_mgga.hpp
@@ -128,16 +128,16 @@ __global__ void eval_vvar_mgga_kern( size_t           ntasks,
       for (int sm_y = threadIdx.y; sm_y < MGGA_KERNEL_SM_BLOCK; sm_y += blockDim.y) {
         const int tid_y = bid_y + sm_y;
 
-        register double tx_reg  = den_shared[0][sm_y][threadIdx.x];
-        register double ty_reg  = den_shared[1][sm_y][threadIdx.x];
-        register double tz_reg  = den_shared[2][sm_y][threadIdx.x];
+        double tx_reg  = den_shared[0][sm_y][threadIdx.x];
+        double ty_reg  = den_shared[1][sm_y][threadIdx.x];
+        double tz_reg  = den_shared[2][sm_y][threadIdx.x];
         // Warp blocks are stored col major
-        register double tau_reg = 0.0;
+        double tau_reg = 0.0;
         tau_reg  = 0.5 * cuda::warp_reduce_sum<warp_size>( tx_reg );
         tau_reg += 0.5 * cuda::warp_reduce_sum<warp_size>( ty_reg );
         tau_reg += 0.5 * cuda::warp_reduce_sum<warp_size>( tz_reg );
 
-        register double lapl_reg = 0.0;
+        double lapl_reg = 0.0;
         if constexpr (need_lapl) {
           lapl_reg = den_shared[3][sm_y][threadIdx.x];
           lapl_reg = cuda::warp_reduce_sum<warp_size>(lapl_reg);