diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl.h
index d7134e557f..1224346e06 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl.h
@@ -228,7 +228,7 @@ template <typename... _Name>
 struct __parallel_for_submitter<__internal::__optional_kernel_name<_Name...>>
 {
     template <typename _ExecutionPolicy, typename _Fp, typename _Index, typename... _Ranges>
-    __future<sycl::event>
+    auto
     operator()(_ExecutionPolicy&& __exec, _Fp __brick, _Index __count, _Ranges&&... __rngs) const
     {
         assert(oneapi::dpl::__ranges::__get_first_range_size(__rngs...) > 0);
@@ -249,7 +249,7 @@ struct __parallel_for_submitter<__internal::__optional_kernel_name<_Name...>>
 //General version of parallel_for, one additional parameter - __count of iterations of loop __cgh.parallel_for,
 //for some algorithms happens that size of processing range is n, but amount of iterations is n/2.
 template <typename _ExecutionPolicy, typename _Fp, typename _Index, typename... _Ranges>
-__future<sycl::event>
+auto
 __parallel_for(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec, _Fp __brick, _Index __count,
                _Ranges&&... __rngs)
 {
@@ -275,7 +275,7 @@ struct __parallel_scan_submitter<_CustomName, __internal::__optional_kernel_name
 {
     template <typename _ExecutionPolicy, typename _Range1, typename _Range2, typename _BinaryOperation,
               typename _InitType, typename _LocalScan, typename _GroupScan, typename _GlobalScan>
-    __future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, typename _InitType::__value_type>>
+    auto
     operator()(_ExecutionPolicy&& __exec, _Range1&& __rng1, _Range2&& __rng2, _BinaryOperation __binary_op,
                _InitType __init, _LocalScan __local_scan, _GroupScan __group_scan, _GlobalScan __global_scan) const
     {
@@ -557,7 +557,7 @@ struct __parallel_copy_if_static_single_group_submitter<_Size, _ElemsPerItem, _W
 {
     template <typename _Policy, typename _InRng, typename _OutRng, typename _InitType, typename _BinaryOperation,
               typename _UnaryOp, typename _Assign>
-    __future<sycl::event, __result_and_scratch_storage<_Policy, _Size>>
+    auto
     operator()(_Policy&& __policy, _InRng&& __in_rng, _OutRng&& __out_rng, ::std::size_t __n, _InitType __init,
                _BinaryOperation __bin_op, _UnaryOp __unary_op, _Assign __assign)
     {
@@ -652,7 +652,7 @@ struct __parallel_copy_if_static_single_group_submitter<_Size, _ElemsPerItem, _W
 
 template <typename _ExecutionPolicy, typename _InRng, typename _OutRng, typename _UnaryOperation, typename _InitType,
           typename _BinaryOperation, typename _Inclusive>
-__future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, typename _InitType::__value_type>>
+auto
 __parallel_transform_scan_single_group(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec,
                                        _InRng&& __in_rng, _OutRng&& __out_rng, ::std::size_t __n,
                                        _UnaryOperation __unary_op, _InitType __init, _BinaryOperation __binary_op,
@@ -746,7 +746,7 @@ __parallel_transform_scan_single_group(oneapi::dpl::__internal::__device_backend
 
 template <typename _ExecutionPolicy, typename _Range1, typename _Range2, typename _BinaryOperation, typename _InitType,
           typename _LocalScan, typename _GroupScan, typename _GlobalScan>
-__future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, typename _InitType::__value_type>>
+auto
 __parallel_transform_scan_base(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec,
                                _Range1&& __in_rng, _Range2&& __out_rng, _BinaryOperation __binary_op, _InitType __init,
                                _LocalScan __local_scan, _GroupScan __group_scan, _GlobalScan __global_scan)
@@ -1812,7 +1812,7 @@ struct __parallel_partial_sort_submitter<__internal::__optional_kernel_name<_Glo
                                          __internal::__optional_kernel_name<_CopyBackName...>>
 {
     template <typename _BackendTag, typename _ExecutionPolicy, typename _Range, typename _Merge, typename _Compare>
-    __future<sycl::event>
+    auto
     operator()(_BackendTag, _ExecutionPolicy&& __exec, _Range&& __rng, _Merge __merge, _Compare __comp) const
     {
         using _Tp = oneapi::dpl::__internal::__value_t<_Range>;
diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_fpga.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_fpga.h
index 3c73eaafdf..2ced9aaec5 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_fpga.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_fpga.h
@@ -56,7 +56,7 @@ template <typename... _Name>
 struct __parallel_for_fpga_submitter<__internal::__optional_kernel_name<_Name...>>
 {
     template <typename _ExecutionPolicy, typename _Fp, typename _Index, typename... _Ranges>
-    __future<sycl::event>
+    auto
     operator()(_ExecutionPolicy&& __exec, _Fp __brick, _Index __count, _Ranges&&... __rngs) const
     {
         auto __n = oneapi::dpl::__ranges::__get_first_range_size(__rngs...);
@@ -80,7 +80,7 @@ struct __parallel_for_fpga_submitter<__internal::__optional_kernel_name<_Name...
 };
 
 template <typename _ExecutionPolicy, typename _Fp, typename _Index, typename... _Ranges>
-__future<sycl::event>
+auto
 __parallel_for(oneapi::dpl::__internal::__fpga_backend_tag, _ExecutionPolicy&& __exec, _Fp __brick, _Index __count,
                _Ranges&&... __rngs)
 {
diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_histogram.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_histogram.h
index dde5cee2d4..f3ba8672f0 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_histogram.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_histogram.h
@@ -497,7 +497,7 @@ __histogram_general_private_global_atomics(oneapi::dpl::__internal::__device_bac
 
 template <::std::uint16_t __iters_per_work_item, typename _ExecutionPolicy, typename _Range1, typename _Range2,
           typename _BinHashMgr>
-__future<sycl::event>
+auto
 __parallel_histogram_select_kernel(oneapi::dpl::__internal::__device_backend_tag __backend_tag,
                                    _ExecutionPolicy&& __exec, const sycl::event& __init_event, _Range1&& __input,
                                    _Range2&& __bins, const _BinHashMgr& __binhash_manager)
diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge.h
index bfdfd8883d..b9990958e9 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge.h
@@ -137,7 +137,7 @@ template <typename _IdType, typename... _Name>
 struct __parallel_merge_submitter<_IdType, __internal::__optional_kernel_name<_Name...>>
 {
     template <typename _ExecutionPolicy, typename _Range1, typename _Range2, typename _Range3, typename _Compare>
-    __future<sycl::event>
+    auto
     operator()(_ExecutionPolicy&& __exec, _Range1&& __rng1, _Range2&& __rng2, _Range3&& __rng3, _Compare __comp) const
     {
         const _IdType __n1 = __rng1.size();
diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge_sort.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge_sort.h
index ef1faf2ef7..1c0bd923a4 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge_sort.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_merge_sort.h
@@ -212,7 +212,7 @@ struct __parallel_sort_submitter<_IdType, __internal::__optional_kernel_name<_Le
                                  __internal::__optional_kernel_name<_CopyBackName...>>
 {
     template <typename _ExecutionPolicy, typename _Range, typename _Compare, typename _LeafSorter>
-    __future<sycl::event>
+    auto
     operator()(_ExecutionPolicy&& __exec, _Range&& __rng, _Compare __comp, _LeafSorter& __leaf_sorter) const
     {
         using _Tp = oneapi::dpl::__internal::__value_t<_Range>;
diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_radix_sort.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_radix_sort.h
index 6d08927ea5..0c11355657 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_radix_sort.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_radix_sort.h
@@ -766,7 +766,7 @@ struct __parallel_radix_sort_iteration
 // radix sort: main function
 //-----------------------------------------------------------------------
 template <bool __is_ascending, typename _Range, typename _ExecutionPolicy, typename _Proj>
-__future<sycl::event>
+auto
 __parallel_radix_sort(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec, _Range&& __in_rng,
                       _Proj __proj)
 {
diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce.h
index 2744251b2f..c0adce402e 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce.h
@@ -125,7 +125,7 @@ struct __parallel_transform_reduce_small_submitter<_Tp, _Commutative, _VecSize,
 {
     template <typename _ExecutionPolicy, typename _Size, typename _ReduceOp, typename _TransformOp, typename _InitType,
               typename... _Ranges>
-    __future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, _Tp>>
+    auto
     operator()(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec, const _Size __n,
                const _Size __work_group_size, const _Size __iters_per_work_item, _ReduceOp __reduce_op,
                _TransformOp __transform_op, _InitType __init, _Ranges&&... __rngs) const
@@ -161,7 +161,7 @@ struct __parallel_transform_reduce_small_submitter<_Tp, _Commutative, _VecSize,
 
 template <typename _Tp, typename _Commutative, std::uint8_t _VecSize, typename _ExecutionPolicy, typename _Size,
           typename _ReduceOp, typename _TransformOp, typename _InitType, typename... _Ranges>
-__future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, _Tp>>
+auto
 __parallel_transform_reduce_small_impl(oneapi::dpl::__internal::__device_backend_tag __backend_tag,
                                        _ExecutionPolicy&& __exec, const _Size __n, const _Size __work_group_size,
                                        const _Size __iters_per_work_item, _ReduceOp __reduce_op,
@@ -236,8 +236,8 @@ struct __parallel_transform_reduce_work_group_kernel_submitter<_Tp, _Commutative
 {
     template <typename _ExecutionPolicy, typename _Size, typename _ReduceOp, typename _InitType,
               typename _ExecutionPolicy2>
-    __future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy2, _Tp>>
-    operator()(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec, sycl::event&& __reduce_event,
+    auto
+    operator()(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec, sycl::event& __reduce_event,
                const _Size __n, const _Size __work_group_size, const _Size __iters_per_work_item, _ReduceOp __reduce_op,
                _InitType __init, __result_and_scratch_storage<_ExecutionPolicy2, _Tp>& __scratch_container) const
     {
@@ -312,7 +312,7 @@ struct __parallel_transform_reduce_impl
 {
     template <typename _ExecutionPolicy, typename _Size, typename _ReduceOp, typename _TransformOp, typename _InitType,
               typename... _Ranges>
-    static __future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, _Tp>>
+    static auto
     submit(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec, _Size __n, _Size __work_group_size,
            const _Size __iters_per_work_item, _ReduceOp __reduce_op, _TransformOp __transform_op, _InitType __init,
            _Ranges&&... __rngs)
@@ -439,7 +439,7 @@ struct __parallel_transform_reduce_impl
 // reduced in each step.
 template <typename _Tp, typename _Commutative, typename _ExecutionPolicy, typename _ReduceOp, typename _TransformOp,
           typename _InitType, typename... _Ranges>
-__future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, _Tp>>
+auto
 __parallel_transform_reduce(oneapi::dpl::__internal::__device_backend_tag __backend_tag, _ExecutionPolicy&& __exec,
                             _ReduceOp __reduce_op, _TransformOp __transform_op, _InitType __init, _Ranges&&... __rngs)
 {
diff --git a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce_then_scan.h b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce_then_scan.h
index a7bd22bdd9..3d69cab952 100644
--- a/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce_then_scan.h
+++ b/include/oneapi/dpl/pstl/hetero/dpcpp/parallel_backend_sycl_reduce_then_scan.h
@@ -747,7 +747,7 @@ __is_gpu_with_sg_32(const _ExecutionPolicy& __exec)
 template <typename _ExecutionPolicy, typename _InRng, typename _OutRng, typename _GenReduceInput, typename _ReduceOp,
           typename _GenScanInput, typename _ScanInputTransform, typename _WriteOp, typename _InitType,
           typename _Inclusive, typename _IsUniquePattern>
-__future<sycl::event, __result_and_scratch_storage<_ExecutionPolicy, typename _InitType::__value_type>>
+auto
 __parallel_transform_reduce_then_scan(oneapi::dpl::__internal::__device_backend_tag, _ExecutionPolicy&& __exec,
                                       _InRng&& __in_rng, _OutRng&& __out_rng, _GenReduceInput __gen_reduce_input,
                                       _ReduceOp __reduce_op, _GenScanInput __gen_scan_input,