diff --git a/level_zero/core/source/kernel/kernel_imp.cpp b/level_zero/core/source/kernel/kernel_imp.cpp
index c7f369fd50..c882b06643 100644
--- a/level_zero/core/source/kernel/kernel_imp.cpp
+++ b/level_zero/core/source/kernel/kernel_imp.cpp
@@ -363,7 +363,7 @@ ze_result_t KernelImp::setGroupSize(uint32_t groupSizeX, uint32_t groupSizeY,
                                          static_cast<uint16_t>(groupSizeY),
                                          static_cast<uint16_t>(groupSizeZ)}},
                 std::array<uint8_t, 3>{{0, 1, 2}},
-                false, grfSize, gfxCoreHelper);
+                false, grfSize);
         }
 
         this->perThreadDataSize = perThreadDataSizeForWholeThreadGroup / numThreadsPerThreadGroup;
diff --git a/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_2.cpp b/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_2.cpp
index 907aadb0fb..c814e6b6b3 100644
--- a/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_2.cpp
+++ b/level_zero/core/test/unit_tests/sources/cmdlist/test_cmdlist_append_launch_kernel_2.cpp
@@ -1023,8 +1023,7 @@ HWCMDTEST_F(IGFX_XE_HP_CORE, CmdlistAppendLaunchKernelWithImplicitArgsTests, giv
 
     auto grfSize = ImplicitArgsHelper::getGrfSize(expectedImplicitArgs.simdWidth);
     auto expectedLocalIds = alignedMalloc(implicitArgsProgrammingSize - sizeof(ImplicitArgs), MemoryConstants::cacheLineSize);
-    const auto &gfxCoreHelper = device->getNEODevice()->getGfxCoreHelper();
-    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, workgroupDimOrder, false, grfSize, gfxCoreHelper);
+    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, workgroupDimOrder, false, grfSize);
 
     auto localIdsProgrammingSize = implicitArgsProgrammingSize - sizeof(ImplicitArgs);
     size_t sizeForLocalIds = NEO::PerThreadDataHelper::getPerThreadDataSizeTotal(expectedImplicitArgs.simdWidth, grfSize, 3u, totalLocalSize);
@@ -1069,8 +1068,7 @@ HWCMDTEST_F(IGFX_XE_HP_CORE, CmdlistAppendLaunchKernelWithImplicitArgsTests, giv
 
     auto grfSize = ImplicitArgsHelper::getGrfSize(expectedImplicitArgs.simdWidth);
     auto expectedLocalIds = alignedMalloc(implicitArgsProgrammingSize - sizeof(ImplicitArgs), MemoryConstants::cacheLineSize);
-    const auto &gfxCoreHelper = device->getNEODevice()->getGfxCoreHelper();
-    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, expectedDimOrder, false, grfSize, gfxCoreHelper);
+    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, expectedDimOrder, false, grfSize);
 
     auto localIdsProgrammingSize = implicitArgsProgrammingSize - sizeof(ImplicitArgs);
     size_t sizeForLocalIds = NEO::PerThreadDataHelper::getPerThreadDataSizeTotal(expectedImplicitArgs.simdWidth, grfSize, 3u, totalLocalSize);
diff --git a/opencl/source/helpers/hardware_commands_helper_bdw_and_later.inl b/opencl/source/helpers/hardware_commands_helper_bdw_and_later.inl
index 44199f2dd0..15c5dab2d9 100644
--- a/opencl/source/helpers/hardware_commands_helper_bdw_and_later.inl
+++ b/opencl/source/helpers/hardware_commands_helper_bdw_and_later.inl
@@ -91,8 +91,7 @@ size_t HardwareCommandsHelper<GfxFamily>::sendCrossThreadData(
 
         auto implicitArgsGpuVA = indirectHeap.getGraphicsAllocation()->getGpuAddress() + indirectHeap.getUsed();
         auto ptrToPatchImplicitArgs = indirectHeap.getSpace(sizeForImplicitArgsProgramming);
-        const auto &gfxCoreHelper = kernel.getGfxCoreHelper();
-        ImplicitArgsHelper::patchImplicitArgs(ptrToPatchImplicitArgs, *pImplicitArgs, kernelDescriptor, {}, gfxCoreHelper);
+        ImplicitArgsHelper::patchImplicitArgs(ptrToPatchImplicitArgs, *pImplicitArgs, kernelDescriptor, {});
 
         auto implicitArgsCrossThreadPtr = ptrOffset(reinterpret_cast<uint64_t *>(kernel.getCrossThreadData()), kernelDescriptor.payloadMappings.implicitArgs.implicitArgsBuffer);
         *implicitArgsCrossThreadPtr = implicitArgsGpuVA;
diff --git a/opencl/source/helpers/hardware_commands_helper_xehp_and_later.inl b/opencl/source/helpers/hardware_commands_helper_xehp_and_later.inl
index 7dd4e52ca4..c463a5a4f9 100644
--- a/opencl/source/helpers/hardware_commands_helper_xehp_and_later.inl
+++ b/opencl/source/helpers/hardware_commands_helper_xehp_and_later.inl
@@ -96,8 +96,8 @@ size_t HardwareCommandsHelper<GfxFamily>::sendCrossThreadData(
             kernelAttributes.flags.requiresWorkgroupWalkOrder,
             requiredWalkOrder,
             kernelDescriptor.kernelAttributes.simdSize);
-        const auto &gfxCoreHelper = kernel.getGfxCoreHelper();
-        ImplicitArgsHelper::patchImplicitArgs(ptrToPatchImplicitArgs, *pImplicitArgs, kernelDescriptor, std::make_pair(generationOfLocalIdsByRuntime, requiredWalkOrder), gfxCoreHelper);
+
+        ImplicitArgsHelper::patchImplicitArgs(ptrToPatchImplicitArgs, *pImplicitArgs, kernelDescriptor, std::make_pair(generationOfLocalIdsByRuntime, requiredWalkOrder));
     }
 
     using InlineData = typename GfxFamily::INLINE_DATA;
diff --git a/opencl/source/kernel/kernel.cpp b/opencl/source/kernel/kernel.cpp
index b9c6f0414a..66961d203d 100644
--- a/opencl/source/kernel/kernel.cpp
+++ b/opencl/source/kernel/kernel.cpp
@@ -581,7 +581,7 @@ cl_int Kernel::getWorkGroupInfo(cl_kernel_work_group_info paramName,
     cl_ulong privateMemSize;
     size_t maxWorkgroupSize;
     const auto &hwInfo = clDevice.getHardwareInfo();
-    auto &gfxCoreHelper = this->getGfxCoreHelper();
+    auto &gfxCoreHelper = clDevice.getGfxCoreHelper();
     auto &clGfxCoreHelper = clDevice.getRootDeviceEnvironment().getHelper<ClGfxCoreHelper>();
     GetInfoHelper info(paramValue, paramValueSize, paramValueSizeRet);
 
@@ -1537,7 +1537,7 @@ cl_int Kernel::setArgBuffer(uint32_t argIndex,
                                    disableL3, isAuxTranslationKernel, arg.isReadOnly(), pClDevice->getDevice(),
                                    kernelInfo.kernelDescriptor.kernelAttributes.flags.useGlobalAtomics, areMultipleSubDevicesInContext());
         } else if (isValidOffset(argAsPtr.bindless)) {
-            auto &gfxCoreHelper = this->getGfxCoreHelper();
+            auto &gfxCoreHelper = getDevice().getGfxCoreHelper();
             auto surfaceStateSize = gfxCoreHelper.getRenderSurfaceStateSize();
             auto surfaceState = ptrOffset(getSurfaceStateHeap(), surfaceStateSize * argIndex);
 
@@ -1655,7 +1655,7 @@ cl_int Kernel::setArgImageWithMipLevel(uint32_t argIndex,
 
         void *surfaceState = nullptr;
         if (isValidOffset(argAsImg.bindless)) {
-            auto &gfxCoreHelper = this->getGfxCoreHelper();
+            auto &gfxCoreHelper = getDevice().getGfxCoreHelper();
             auto surfaceStateSize = gfxCoreHelper.getRenderSurfaceStateSize();
             surfaceState = ptrOffset(getSurfaceStateHeap(), surfaceStateSize * argIndex);
         } else {
@@ -2130,7 +2130,7 @@ uint64_t Kernel::getKernelStartAddress(const bool localIdsGenerationByRuntime, c
     kernelStartOffset += getStartOffset();
 
     auto &hardwareInfo = getHardwareInfo();
-    const auto &gfxCoreHelper = this->getGfxCoreHelper();
+    const auto &gfxCoreHelper = getDevice().getGfxCoreHelper();
     const auto &productHelper = getDevice().getProductHelper();
 
     if (isCssUsed && gfxCoreHelper.isOffsetToSkipSetFFIDGPWARequired(hardwareInfo, productHelper)) {
@@ -2140,7 +2140,7 @@ uint64_t Kernel::getKernelStartAddress(const bool localIdsGenerationByRuntime, c
     return kernelStartOffset;
 }
 void *Kernel::patchBindlessSurfaceState(NEO::GraphicsAllocation *alloc, uint32_t bindless) {
-    auto &gfxCoreHelper = this->getGfxCoreHelper();
+    auto &gfxCoreHelper = getDevice().getGfxCoreHelper();
     auto surfaceStateSize = gfxCoreHelper.getRenderSurfaceStateSize();
     NEO::BindlessHeapsHelper *bindlessHeapsHelper = getDevice().getDevice().getBindlessHeapsHelper();
     auto ssInHeap = bindlessHeapsHelper->allocateSSInHeap(surfaceStateSize, alloc, NEO::BindlessHeapsHelper::GLOBAL_SSH);
@@ -2159,7 +2159,7 @@ uint32_t Kernel::getAdditionalKernelExecInfo() const {
 }
 
 bool Kernel::requiresWaDisableRccRhwoOptimization() const {
-    auto &gfxCoreHelper = this->getGfxCoreHelper();
+    auto &gfxCoreHelper = getDevice().getGfxCoreHelper();
     auto rootDeviceIndex = getDevice().getRootDeviceIndex();
 
     if (gfxCoreHelper.isWaDisableRccRhwoOptimizationRequired() && isUsingSharedObjArgs()) {
@@ -2270,7 +2270,7 @@ void Kernel::reconfigureKernel() {
         kernelDescriptor.kernelAttributes.simdSize != 32) {
         this->maxKernelWorkGroupSize >>= 1;
     }
-    const auto &gfxCoreHelper = this->getGfxCoreHelper();
+    const auto &gfxCoreHelper = getDevice().getGfxCoreHelper();
     bool isLocalIdsGeneratedByHw = false; // if local ids generated by runtime then more work groups available
     auto maxWorkGroupSize = static_cast<uint32_t>(kernelInfo.getMaxRequiredWorkGroupSize(getMaxKernelWorkGroupSize()));
 
@@ -2369,8 +2369,7 @@ void Kernel::initializeLocalIdsCache() {
 
 void Kernel::setLocalIdsForGroup(const Vec3<uint16_t> &groupSize, void *destination) const {
     UNRECOVERABLE_IF(localIdsCache.get() == nullptr);
-    const auto &gfxCoreHelper = this->getGfxCoreHelper();
-    localIdsCache->setLocalIdsForGroup(groupSize, destination, gfxCoreHelper);
+    localIdsCache->setLocalIdsForGroup(groupSize, destination);
 }
 
 size_t Kernel::getLocalIdsSizeForGroup(const Vec3<uint16_t> &groupSize) const {
diff --git a/opencl/source/kernel/kernel.h b/opencl/source/kernel/kernel.h
index b86ba4d0c5..94f100c8fc 100644
--- a/opencl/source/kernel/kernel.h
+++ b/opencl/source/kernel/kernel.h
@@ -410,10 +410,6 @@ class Kernel : public ReferenceTrackedObject<Kernel> {
     size_t getLocalIdsSizeForGroup(const Vec3<uint16_t> &groupSize) const;
     size_t getLocalIdsSizePerThread() const;
 
-    const GfxCoreHelper &getGfxCoreHelper() const {
-        return getDevice().getGfxCoreHelper();
-    }
-
   protected:
     struct KernelConfig {
         Vec3<size_t> gws;
diff --git a/opencl/test/unit_test/helpers/hardware_commands_helper_tests.cpp b/opencl/test/unit_test/helpers/hardware_commands_helper_tests.cpp
index 76dbfdd63c..7afbccf43f 100644
--- a/opencl/test/unit_test/helpers/hardware_commands_helper_tests.cpp
+++ b/opencl/test/unit_test/helpers/hardware_commands_helper_tests.cpp
@@ -552,13 +552,12 @@ HWCMDTEST_F(IGFX_GEN8_CORE, HardwareCommandsTest, whenSendingIndirectStateThenKe
     ASSERT_LE(expectedIohSize, ioh.getUsed());
 
     auto expectedLocalIds = alignedMalloc(expectedIohSize, 64);
-    const auto &gfxCoreHelper = pDevice->getGfxCoreHelper();
     generateLocalIDs(expectedLocalIds, modifiedKernelInfo.getMaxSimdSize(),
                      std::array<uint16_t, 3>{{localWorkSizeX, localWorkSizeY, localWorkSizeZ}},
                      std::array<uint8_t, 3>{{modifiedKernelInfo.kernelDescriptor.kernelAttributes.workgroupDimensionsOrder[0],
                                              modifiedKernelInfo.kernelDescriptor.kernelAttributes.workgroupDimensionsOrder[1],
                                              modifiedKernelInfo.kernelDescriptor.kernelAttributes.workgroupDimensionsOrder[2]}},
-                     false, grfSize, gfxCoreHelper);
+                     false, grfSize);
 
     EXPECT_EQ(0, memcmp(expectedLocalIds, ioh.getCpuBase(), expectedIohSize));
     alignedFree(expectedLocalIds);
@@ -1378,8 +1377,7 @@ HWCMDTEST_F(IGFX_XE_HP_CORE, HardwareCommandsImplicitArgsTests, givenKernelWithI
 
     auto grfSize = ImplicitArgsHelper::getGrfSize(expectedImplicitArgs.simdWidth);
     auto expectedLocalIds = alignedMalloc(implicitArgsProgrammingSize - sizeof(ImplicitArgs), MemoryConstants::cacheLineSize);
-    const auto &gfxCoreHelper = pDevice->getGfxCoreHelper();
-    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, workgroupDimOrder, false, grfSize, gfxCoreHelper);
+    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, workgroupDimOrder, false, grfSize);
 
     auto localIdsProgrammingSize = implicitArgsProgrammingSize - sizeof(ImplicitArgs);
     size_t sizeForLocalIds = PerThreadDataHelper::getPerThreadDataSizeTotal(expectedImplicitArgs.simdWidth, grfSize, 3u, totalLocalSize);
@@ -1412,8 +1410,7 @@ HWCMDTEST_F(IGFX_XE_HP_CORE, HardwareCommandsImplicitArgsTests, givenKernelWithI
 
     auto grfSize = ImplicitArgsHelper::getGrfSize(expectedImplicitArgs.simdWidth);
     auto expectedLocalIds = alignedMalloc(implicitArgsProgrammingSize - sizeof(ImplicitArgs), MemoryConstants::cacheLineSize);
-    const auto &gfxCoreHelper = pDevice->getGfxCoreHelper();
-    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, expectedDimOrder, false, grfSize, gfxCoreHelper);
+    generateLocalIDs(expectedLocalIds, expectedImplicitArgs.simdWidth, localSize, expectedDimOrder, false, grfSize);
 
     auto localIdsProgrammingSize = implicitArgsProgrammingSize - sizeof(ImplicitArgs);
     size_t sizeForLocalIds = PerThreadDataHelper::getPerThreadDataSizeTotal(expectedImplicitArgs.simdWidth, grfSize, 3u, totalLocalSize);
diff --git a/shared/source/command_container/command_encoder_bdw_and_later.inl b/shared/source/command_container/command_encoder_bdw_and_later.inl
index 19c7e9eef6..0722f99a9b 100644
--- a/shared/source/command_container/command_encoder_bdw_and_later.inl
+++ b/shared/source/command_container/command_encoder_bdw_and_later.inl
@@ -183,7 +183,7 @@ void EncodeDispatchKernel<Family>::encode(CommandContainer &container, EncodeDis
             auto implicitArgsCrossThreadPtr = ptrOffset(const_cast<uint64_t *>(reinterpret_cast<const uint64_t *>(args.dispatchInterface->getCrossThreadData())), kernelDescriptor.payloadMappings.implicitArgs.implicitArgsBuffer);
             *implicitArgsCrossThreadPtr = implicitArgsGpuVA;
 
-            ptr = NEO::ImplicitArgsHelper::patchImplicitArgs(ptr, *pImplicitArgs, kernelDescriptor, {}, gfxCoreHelper);
+            ptr = NEO::ImplicitArgsHelper::patchImplicitArgs(ptr, *pImplicitArgs, kernelDescriptor, {});
         }
 
         memcpy_s(ptr, sizeCrossThreadData,
diff --git a/shared/source/command_container/command_encoder_xehp_and_later.inl b/shared/source/command_container/command_encoder_xehp_and_later.inl
index a14be7c67f..d6e711509e 100644
--- a/shared/source/command_container/command_encoder_xehp_and_later.inl
+++ b/shared/source/command_container/command_encoder_xehp_and_later.inl
@@ -218,7 +218,7 @@ void EncodeDispatchKernel<Family>::encode(CommandContainer &container, EncodeDis
         if (pImplicitArgs) {
             offsetThreadData -= sizeof(ImplicitArgs);
             pImplicitArgs->localIdTablePtr = heap->getGraphicsAllocation()->getGpuAddress() + heap->getUsed() - iohRequiredSize;
-            ptr = NEO::ImplicitArgsHelper::patchImplicitArgs(ptr, *pImplicitArgs, kernelDescriptor, std::make_pair(localIdsGenerationByRuntime, requiredWorkgroupOrder), gfxCoreHelper);
+            ptr = NEO::ImplicitArgsHelper::patchImplicitArgs(ptr, *pImplicitArgs, kernelDescriptor, std::make_pair(localIdsGenerationByRuntime, requiredWorkgroupOrder));
         }
 
         if (sizeCrossThreadData > 0) {
diff --git a/shared/source/helpers/aarch64/local_id_gen.cpp b/shared/source/helpers/aarch64/local_id_gen.cpp
index 4169785bf7..8a3cd04ea7 100644
--- a/shared/source/helpers/aarch64/local_id_gen.cpp
+++ b/shared/source/helpers/aarch64/local_id_gen.cpp
@@ -8,7 +8,6 @@
 #include "shared/source/helpers/local_id_gen.h"
 
 #include "shared/source/helpers/aligned_memory.h"
-#include "shared/source/helpers/gfx_core_helper.h"
 #include "shared/source/helpers/local_id_gen_special.inl"
 #include "shared/source/utilities/cpu_info.h"
 
@@ -42,9 +41,8 @@ LocalIDHelper::LocalIDHelper() {
 
 LocalIDHelper LocalIDHelper::initializer;
 
-void generateLocalIDs(void *buffer, uint16_t simd, const std::array<uint16_t, 3> &localWorkgroupSize, const std::array<uint8_t, 3> &dimensionsOrder, bool isImageOnlyKernel, uint32_t grfSize, const GfxCoreHelper &gfxCoreHelper) {
-    bool localIdsGeneratedByHw = false;
-    auto threadsPerWorkGroup = static_cast<uint16_t>(gfxCoreHelper.calculateNumThreadsPerThreadGroup(simd, static_cast<uint32_t>(localWorkgroupSize[0] * localWorkgroupSize[1] * localWorkgroupSize[2]), grfSize, localIdsGeneratedByHw));
+void generateLocalIDs(void *buffer, uint16_t simd, const std::array<uint16_t, 3> &localWorkgroupSize, const std::array<uint8_t, 3> &dimensionsOrder, bool isImageOnlyKernel, uint32_t grfSize) {
+    auto threadsPerWorkGroup = static_cast<uint16_t>(getThreadsPerWG(simd, static_cast<uint32_t>(localWorkgroupSize[0] * localWorkgroupSize[1] * localWorkgroupSize[2])));
     bool useLayoutForImages = isImageOnlyKernel && isCompatibleWithLayoutForImages(localWorkgroupSize, dimensionsOrder, simd);
     if (useLayoutForImages) {
         generateLocalIDsWithLayoutForImages(buffer, localWorkgroupSize, simd);
diff --git a/shared/source/helpers/local_id_gen.h b/shared/source/helpers/local_id_gen.h
index ab88965571..94fb33dd5e 100644
--- a/shared/source/helpers/local_id_gen.h
+++ b/shared/source/helpers/local_id_gen.h
@@ -12,7 +12,6 @@
 #include <cstdint>
 
 namespace NEO {
-class GfxCoreHelper;
 inline uint32_t getGRFsPerThread(uint32_t simd, uint32_t grfSize) {
     return (simd == 32 && grfSize == 32) ? 2 : 1;
 }
@@ -59,7 +58,7 @@ void generateLocalIDsSimd(void *b, const std::array<uint16_t, 3> &localWorkgroup
                           const std::array<uint8_t, 3> &dimensionsOrder, bool chooseMaxRowSize);
 
 void generateLocalIDs(void *buffer, uint16_t simd, const std::array<uint16_t, 3> &localWorkgroupSize,
-                      const std::array<uint8_t, 3> &dimensionsOrder, bool isImageOnlyKernel, uint32_t grfSize, const NEO::GfxCoreHelper &gfxCoreHelper);
+                      const std::array<uint8_t, 3> &dimensionsOrder, bool isImageOnlyKernel, uint32_t grfSize);
 void generateLocalIDsWithLayoutForImages(void *b, const std::array<uint16_t, 3> &localWorkgroupSize, uint16_t simd);
 
 bool isCompatibleWithLayoutForImages(const std::array<uint16_t, 3> &localWorkgroupSize, const std::array<uint8_t, 3> &dimensionsOrder, uint16_t simd);
diff --git a/shared/source/helpers/x86_64/local_id_gen.cpp b/shared/source/helpers/x86_64/local_id_gen.cpp
index 3c7a9def07..18c464284f 100644
--- a/shared/source/helpers/x86_64/local_id_gen.cpp
+++ b/shared/source/helpers/x86_64/local_id_gen.cpp
@@ -8,7 +8,6 @@
 #include "shared/source/helpers/local_id_gen.h"
 
 #include "shared/source/helpers/aligned_memory.h"
-#include "shared/source/helpers/gfx_core_helper.h"
 #include "shared/source/helpers/local_id_gen_special.inl"
 #include "shared/source/utilities/cpu_info.h"
 
@@ -45,9 +44,8 @@ LocalIDHelper::LocalIDHelper() {
 LocalIDHelper LocalIDHelper::initializer;
 
 // traditional function to generate local IDs
-void generateLocalIDs(void *buffer, uint16_t simd, const std::array<uint16_t, 3> &localWorkgroupSize, const std::array<uint8_t, 3> &dimensionsOrder, bool isImageOnlyKernel, uint32_t grfSize, const GfxCoreHelper &gfxCoreHelper) {
-    bool localIdsGeneratedByHw = false;
-    auto threadsPerWorkGroup = static_cast<uint16_t>(gfxCoreHelper.calculateNumThreadsPerThreadGroup(simd, static_cast<uint32_t>(localWorkgroupSize[0] * localWorkgroupSize[1] * localWorkgroupSize[2]), grfSize, localIdsGeneratedByHw));
+void generateLocalIDs(void *buffer, uint16_t simd, const std::array<uint16_t, 3> &localWorkgroupSize, const std::array<uint8_t, 3> &dimensionsOrder, bool isImageOnlyKernel, uint32_t grfSize) {
+    auto threadsPerWorkGroup = static_cast<uint16_t>(getThreadsPerWG(simd, static_cast<uint32_t>(localWorkgroupSize[0] * localWorkgroupSize[1] * localWorkgroupSize[2])));
     bool useLayoutForImages = isImageOnlyKernel && isCompatibleWithLayoutForImages(localWorkgroupSize, dimensionsOrder, simd);
     if (useLayoutForImages) {
         generateLocalIDsWithLayoutForImages(buffer, localWorkgroupSize, simd);
diff --git a/shared/source/kernel/implicit_args.h b/shared/source/kernel/implicit_args.h
index 9046809586..1eb6090cc7 100644
--- a/shared/source/kernel/implicit_args.h
+++ b/shared/source/kernel/implicit_args.h
@@ -15,7 +15,6 @@
 namespace NEO {
 
 struct KernelDescriptor;
-class GfxCoreHelper;
 
 struct ImplicitArgs {
     uint8_t structSize;
@@ -52,6 +51,6 @@ namespace ImplicitArgsHelper {
 std::array<uint8_t, 3> getDimensionOrderForLocalIds(const uint8_t *workgroupDimensionsOrder, std::optional<std::pair<bool /* localIdsGeneratedByRuntime */, uint32_t /* walkOrderForHwGenerationOfLocalIds */>> hwGenerationOfLocalIdsParams);
 uint32_t getGrfSize(uint32_t simd);
 uint32_t getSizeForImplicitArgsPatching(const ImplicitArgs *pImplicitArgs, const KernelDescriptor &kernelDescriptor);
-void *patchImplicitArgs(void *ptrToPatch, const ImplicitArgs &implicitArgs, const KernelDescriptor &kernelDescriptor, std::optional<std::pair<bool /* localIdsGeneratedByRuntime */, uint32_t /* walkOrderForHwGenerationOfLocalIds */>> hwGenerationOfLocalIdsParams, const GfxCoreHelper &gfxCoreHelper);
+void *patchImplicitArgs(void *ptrToPatch, const ImplicitArgs &implicitArgs, const KernelDescriptor &kernelDescriptor, std::optional<std::pair<bool /* localIdsGeneratedByRuntime */, uint32_t /* walkOrderForHwGenerationOfLocalIds */>> hwGenerationOfLocalIdsParams);
 } // namespace ImplicitArgsHelper
 } // namespace NEO
diff --git a/shared/source/kernel/implicit_args_helper.cpp b/shared/source/kernel/implicit_args_helper.cpp
index 7b0b7ecdc8..24f783a705 100644
--- a/shared/source/kernel/implicit_args_helper.cpp
+++ b/shared/source/kernel/implicit_args_helper.cpp
@@ -62,7 +62,7 @@ uint32_t getSizeForImplicitArgsPatching(const ImplicitArgs *pImplicitArgs, const
     }
 }
 
-void *patchImplicitArgs(void *ptrToPatch, const ImplicitArgs &implicitArgs, const KernelDescriptor &kernelDescriptor, std::optional<std::pair<bool, uint32_t>> hwGenerationOfLocalIdsParams, const GfxCoreHelper &gfxCoreHelper) {
+void *patchImplicitArgs(void *ptrToPatch, const ImplicitArgs &implicitArgs, const KernelDescriptor &kernelDescriptor, std::optional<std::pair<bool, uint32_t>> hwGenerationOfLocalIdsParams) {
 
     auto totalSizeToProgram = getSizeForImplicitArgsPatching(&implicitArgs, kernelDescriptor);
     auto retVal = ptrOffset(ptrToPatch, totalSizeToProgram);
@@ -80,7 +80,7 @@ void *patchImplicitArgs(void *ptrToPatch, const ImplicitArgs &implicitArgs, cons
                                      static_cast<uint16_t>(implicitArgs.localSizeY),
                                      static_cast<uint16_t>(implicitArgs.localSizeZ)}},
             dimensionOrder,
-            false, grfSize, gfxCoreHelper);
+            false, grfSize);
         auto sizeForLocalIdsProgramming = totalSizeToProgram - sizeof(NEO::ImplicitArgs);
         ptrToPatch = ptrOffset(ptrToPatch, sizeForLocalIdsProgramming);
     }
diff --git a/shared/source/kernel/local_ids_cache.cpp b/shared/source/kernel/local_ids_cache.cpp
index f678b72fbb..6fb149297a 100644
--- a/shared/source/kernel/local_ids_cache.cpp
+++ b/shared/source/kernel/local_ids_cache.cpp
@@ -47,7 +47,7 @@ void LocalIdsCache::setLocalIdsForEntry(LocalIdsCacheEntry &entry, void *destina
     std::memcpy(destination, entry.localIdsData, entry.localIdsSize);
 }
 
-void LocalIdsCache::setLocalIdsForGroup(const Vec3<uint16_t> &group, void *destination, const GfxCoreHelper &gfxCoreHelper) {
+void LocalIdsCache::setLocalIdsForGroup(const Vec3<uint16_t> &group, void *destination) {
     auto setLocalIdsLock = lock();
     LocalIdsCacheEntry *leastAccessedEntry = &cache[0];
     for (auto &cacheEntry : cache) {
@@ -60,11 +60,11 @@ void LocalIdsCache::setLocalIdsForGroup(const Vec3<uint16_t> &group, void *desti
         }
     }
 
-    commitNewEntry(*leastAccessedEntry, group, gfxCoreHelper);
+    commitNewEntry(*leastAccessedEntry, group);
     setLocalIdsForEntry(*leastAccessedEntry, destination);
 }
 
-void LocalIdsCache::commitNewEntry(LocalIdsCacheEntry &entry, const Vec3<uint16_t> &group, const GfxCoreHelper &gfxCoreHelper) {
+void LocalIdsCache::commitNewEntry(LocalIdsCacheEntry &entry, const Vec3<uint16_t> &group) {
     entry.localIdsSize = getLocalIdsSizeForGroup(group);
     entry.groupSize = group;
     entry.accessCounter = 0U;
@@ -74,7 +74,7 @@ void LocalIdsCache::commitNewEntry(LocalIdsCacheEntry &entry, const Vec3<uint16_
         entry.localIdsSizeAllocated = entry.localIdsSize;
     }
     NEO::generateLocalIDs(entry.localIdsData, static_cast<uint16_t>(simdSize),
-                          {group[0], group[1], group[2]}, wgDimOrder, usesOnlyImages, grfSize, gfxCoreHelper);
+                          {group[0], group[1], group[2]}, wgDimOrder, usesOnlyImages, grfSize);
 }
 
 } // namespace NEO
\ No newline at end of file
diff --git a/shared/source/kernel/local_ids_cache.h b/shared/source/kernel/local_ids_cache.h
index 26955ed6b0..f7fb37d262 100644
--- a/shared/source/kernel/local_ids_cache.h
+++ b/shared/source/kernel/local_ids_cache.h
@@ -12,7 +12,7 @@
 #include <mutex>
 
 namespace NEO {
-class GfxCoreHelper;
+
 class LocalIdsCache {
   public:
     struct LocalIdsCacheEntry {
@@ -30,13 +30,13 @@ class LocalIdsCache {
     LocalIdsCache(size_t cacheSize, std::array<uint8_t, 3> wgDimOrder, uint8_t simdSize, uint8_t grfSize, bool usesOnlyImages = false);
     ~LocalIdsCache();
 
-    void setLocalIdsForGroup(const Vec3<uint16_t> &group, void *destination, const GfxCoreHelper &gfxCoreHelper);
+    void setLocalIdsForGroup(const Vec3<uint16_t> &group, void *destination);
     size_t getLocalIdsSizeForGroup(const Vec3<uint16_t> &group) const;
     size_t getLocalIdsSizePerThread() const;
 
   protected:
     void setLocalIdsForEntry(LocalIdsCacheEntry &entry, void *destination);
-    void commitNewEntry(LocalIdsCacheEntry &entry, const Vec3<uint16_t> &group, const GfxCoreHelper &gfxCoreHelper);
+    void commitNewEntry(LocalIdsCacheEntry &entry, const Vec3<uint16_t> &group);
     std::unique_lock<std::mutex> lock();
 
     StackVec<LocalIdsCacheEntry, 4> cache;
diff --git a/shared/test/unit_test/helpers/local_id_tests.cpp b/shared/test/unit_test/helpers/local_id_tests.cpp
index 661277ce9b..4cbf58c52e 100644
--- a/shared/test/unit_test/helpers/local_id_tests.cpp
+++ b/shared/test/unit_test/helpers/local_id_tests.cpp
@@ -7,10 +7,8 @@
 
 #include "shared/source/helpers/aligned_memory.h"
 #include "shared/source/helpers/basic_math.h"
-#include "shared/source/helpers/gfx_core_helper.h"
 #include "shared/source/helpers/local_id_gen.h"
 #include "shared/source/helpers/ptr_math.h"
-#include "shared/test/common/helpers/default_hw_info.h"
 #include "shared/test/common/helpers/unit_test_helper.h"
 #include "shared/test/common/test_macros/hw_test.h"
 
@@ -77,16 +75,14 @@ TEST(LocalID, GivenSimd1WhenGettingPerThreadSizeLocalIdsThenValueIsEqualGrfSize)
 
     EXPECT_EQ(grfSize, getPerThreadSizeLocalIDs(simd, grfSize));
 }
-
-TEST(LocalIdTest, givenVariadicGrfSizeWhenLocalSizesAreEmittedThenUseFullRowSize) {
+TEST(LocalID, givenVariadicGrfSizeWhenLocalSizesAreEmittedThenUseFullRowSize) {
     auto localIdsPtr = allocateAlignedMemory(3 * 64u, MemoryConstants::cacheLineSize);
 
     uint16_t *localIdsView = reinterpret_cast<uint16_t *>(localIdsPtr.get());
     std::array<uint16_t, 3u> localSizes = {{2u, 2u, 1u}};
     std::array<uint8_t, 3u> dimensionsOrder = {{0u, 1u, 2u}};
 
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
-    generateLocalIDs(localIdsPtr.get(), 16u, localSizes, dimensionsOrder, false, 64u, *gfxCoreHelper.get());
+    generateLocalIDs(localIdsPtr.get(), 16u, localSizes, dimensionsOrder, false, 64u);
     EXPECT_EQ(localIdsView[0], 0u);
     EXPECT_EQ(localIdsView[1], 1u);
     EXPECT_EQ(localIdsView[2], 0u);
@@ -281,42 +277,37 @@ struct LocalIDFixture : ::testing::TestWithParam<std::tuple<int, int, int, int,
 };
 
 HWTEST_P(LocalIDFixture, WhenGeneratingLocalIdsThenIdsAreWithinLimits) {
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
     generateLocalIDs(buffer, simd, std::array<uint16_t, 3>{{static_cast<uint16_t>(localWorkSizeX), static_cast<uint16_t>(localWorkSizeY), static_cast<uint16_t>(localWorkSizeZ)}},
-                     std::array<uint8_t, 3>{{0, 1, 2}}, false, grfSize, *gfxCoreHelper.get());
+                     std::array<uint8_t, 3>{{0, 1, 2}}, false, grfSize);
     validateIDWithinLimits(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, UnitTestHelper<FamilyType>::useFullRowForLocalIdsGeneration);
 }
 
 HWTEST_P(LocalIDFixture, WhenGeneratingLocalIdsThenAllWorkItemsCovered) {
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
     generateLocalIDs(buffer, simd, std::array<uint16_t, 3>{{static_cast<uint16_t>(localWorkSizeX), static_cast<uint16_t>(localWorkSizeY), static_cast<uint16_t>(localWorkSizeZ)}},
-                     std::array<uint8_t, 3>{{0, 1, 2}}, false, grfSize, *gfxCoreHelper.get());
+                     std::array<uint8_t, 3>{{0, 1, 2}}, false, grfSize);
     validateAllWorkItemsCovered(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, UnitTestHelper<FamilyType>::useFullRowForLocalIdsGeneration);
 }
 
 HWTEST_P(LocalIDFixture, WhenWalkOrderIsXyzThenProperLocalIdsAreGenerated) {
     auto dimensionsOrder = std::array<uint8_t, 3>{{0, 1, 2}};
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
     generateLocalIDs(buffer, simd, std::array<uint16_t, 3>{{static_cast<uint16_t>(localWorkSizeX), static_cast<uint16_t>(localWorkSizeY), static_cast<uint16_t>(localWorkSizeZ)}},
-                     dimensionsOrder, false, grfSize, *gfxCoreHelper.get());
+                     dimensionsOrder, false, grfSize);
     validateAllWorkItemsCovered(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, UnitTestHelper<FamilyType>::useFullRowForLocalIdsGeneration);
     validateWalkOrder(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, dimensionsOrder);
 }
 
 HWTEST_P(LocalIDFixture, WhenWalkOrderIsYxzThenProperLocalIdsAreGenerated) {
     auto dimensionsOrder = std::array<uint8_t, 3>{{1, 0, 2}};
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
     generateLocalIDs(buffer, simd, std::array<uint16_t, 3>{{static_cast<uint16_t>(localWorkSizeX), static_cast<uint16_t>(localWorkSizeY), static_cast<uint16_t>(localWorkSizeZ)}},
-                     dimensionsOrder, false, grfSize, *gfxCoreHelper.get());
+                     dimensionsOrder, false, grfSize);
     validateAllWorkItemsCovered(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, UnitTestHelper<FamilyType>::useFullRowForLocalIdsGeneration);
     validateWalkOrder(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, dimensionsOrder);
 }
 
 HWTEST_P(LocalIDFixture, WhenWalkOrderIsZyxThenProperLocalIdsAreGenerated) {
     auto dimensionsOrder = std::array<uint8_t, 3>{{2, 1, 0}};
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
     generateLocalIDs(buffer, simd, std::array<uint16_t, 3>{{static_cast<uint16_t>(localWorkSizeX), static_cast<uint16_t>(localWorkSizeY), static_cast<uint16_t>(localWorkSizeZ)}},
-                     dimensionsOrder, false, grfSize, *gfxCoreHelper.get());
+                     dimensionsOrder, false, grfSize);
     validateAllWorkItemsCovered(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, UnitTestHelper<FamilyType>::useFullRowForLocalIdsGeneration);
     validateWalkOrder(simd, localWorkSizeX, localWorkSizeY, localWorkSizeZ, dimensionsOrder);
 }
@@ -344,8 +335,8 @@ struct LocalIdsLayoutForImagesTest : ::testing::TestWithParam<std::tuple<uint16_
         rowWidth = simd == 32u ? 32u : 16u;
         xDelta = simd == 8u ? 2u : 4u;
     }
-
     void generateLocalIds() {
+
         auto numGrfs = (localWorkSize.at(0) * localWorkSize.at(1) + (simd - 1)) / simd;
         elemsInBuffer = 3u * simd * numGrfs;
         if (simd == 8u) {
@@ -356,8 +347,7 @@ struct LocalIdsLayoutForImagesTest : ::testing::TestWithParam<std::tuple<uint16_
         memset(memory.get(), 0xff, size);
         buffer = reinterpret_cast<uint16_t *>(memory.get());
         EXPECT_TRUE(isCompatibleWithLayoutForImages(localWorkSize, dimensionsOrder, simd));
-        auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
-        generateLocalIDs(buffer, simd, localWorkSize, dimensionsOrder, true, grfSize, *gfxCoreHelper.get());
+        generateLocalIDs(buffer, simd, localWorkSize, dimensionsOrder, true, grfSize);
     }
     void validateGRF() {
         uint32_t totalLocalIds = localWorkSize.at(0) * localWorkSize.at(1);
@@ -457,9 +447,9 @@ TEST_P(LocalIdsLayoutTest, givenLocalWorkgroupSize4x4x1WhenGenerateLocalIdsThenH
     auto alignedMemory2 = allocateAlignedMemory(size, 32);
     auto buffer2 = reinterpret_cast<uint16_t *>(alignedMemory2.get());
     memset(buffer2, 0xff, size);
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
-    generateLocalIDs(buffer1, simd, localWorkSize, dimensionsOrder, false, grfSize, *gfxCoreHelper.get());
-    generateLocalIDs(buffer2, simd, localWorkSize, dimensionsOrder, true, grfSize, *gfxCoreHelper.get());
+
+    generateLocalIDs(buffer1, simd, localWorkSize, dimensionsOrder, false, grfSize);
+    generateLocalIDs(buffer2, simd, localWorkSize, dimensionsOrder, true, grfSize);
 
     for (auto i = 0u; i < elemsInBuffer / rowWidth; i++) {
         for (auto j = 0u; j < rowWidth; j++) {
diff --git a/shared/test/unit_test/kernel/implicit_args_helper_tests.cpp b/shared/test/unit_test/kernel/implicit_args_helper_tests.cpp
index e5401b893d..aefe5d7590 100644
--- a/shared/test/unit_test/kernel/implicit_args_helper_tests.cpp
+++ b/shared/test/unit_test/kernel/implicit_args_helper_tests.cpp
@@ -6,13 +6,11 @@
  */
 
 #include "shared/source/helpers/aligned_memory.h"
-#include "shared/source/helpers/gfx_core_helper.h"
 #include "shared/source/helpers/hw_walk_order.h"
 #include "shared/source/helpers/per_thread_data.h"
 #include "shared/source/helpers/ptr_math.h"
 #include "shared/source/kernel/implicit_args.h"
 #include "shared/source/kernel/kernel_descriptor.h"
-#include "shared/test/common/helpers/default_hw_info.h"
 #include "shared/test/common/test_macros/hw_test.h"
 
 using namespace NEO;
@@ -119,8 +117,8 @@ TEST(ImplicitArgsHelperTest, givenImplicitArgsWithoutImplicitArgsBufferOffsetInP
     uint8_t pattern = 0xcd;
 
     memset(memoryToPatch.get(), pattern, totalSizeForPatching);
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
-    auto retVal = ImplicitArgsHelper::patchImplicitArgs(memoryToPatch.get(), implicitArgs, kernelDescriptor, {}, *gfxCoreHelper.get());
+
+    auto retVal = ImplicitArgsHelper::patchImplicitArgs(memoryToPatch.get(), implicitArgs, kernelDescriptor, {});
 
     EXPECT_EQ(retVal, ptrOffset(memoryToPatch.get(), totalSizeForPatching));
 
@@ -160,8 +158,8 @@ TEST(ImplicitArgsHelperTest, givenImplicitArgsWithImplicitArgsBufferOffsetInPayl
     uint8_t pattern = 0xcd;
 
     memset(memoryToPatch.get(), pattern, totalSizeForPatching);
-    auto gfxCoreHelper = GfxCoreHelper::create(defaultHwInfo->platform.eRenderCoreFamily);
-    auto retVal = ImplicitArgsHelper::patchImplicitArgs(memoryToPatch.get(), implicitArgs, kernelDescriptor, {}, *gfxCoreHelper.get());
+
+    auto retVal = ImplicitArgsHelper::patchImplicitArgs(memoryToPatch.get(), implicitArgs, kernelDescriptor, {});
 
     EXPECT_EQ(retVal, ptrOffset(memoryToPatch.get(), totalSizeForPatching));
 
diff --git a/shared/test/unit_test/kernel/local_ids_cache_tests.cpp b/shared/test/unit_test/kernel/local_ids_cache_tests.cpp
index 602832f91b..89f585ddb8 100644
--- a/shared/test/unit_test/kernel/local_ids_cache_tests.cpp
+++ b/shared/test/unit_test/kernel/local_ids_cache_tests.cpp
@@ -7,11 +7,8 @@
 
 #include "shared/source/command_stream/linear_stream.h"
 #include "shared/source/helpers/aligned_memory.h"
-#include "shared/source/helpers/gfx_core_helper.h"
-#include "shared/source/helpers/hw_info.h"
 #include "shared/source/helpers/per_thread_data.h"
 #include "shared/source/kernel/local_ids_cache.h"
-#include "shared/test/common/helpers/default_hw_info.h"
 #include "shared/test/common/mocks/mock_graphics_allocation.h"
 #include "shared/test/common/test_macros/test.h"
 
@@ -38,8 +35,7 @@ using LocalIdsCacheTest = Test<LocalIdsCacheFixture>;
 TEST_F(LocalIdsCacheTest, GivenCacheMissWhenGetLocalIdsForGroupThenNewEntryIsCommitedIntoLeastUsedEntry) {
     localIdsCache->cache.resize(2);
     localIdsCache->cache[0].accessCounter = 2U;
-    auto gfxCoreHelper = NEO::GfxCoreHelper::create(NEO::defaultHwInfo->platform.eRenderCoreFamily);
-    localIdsCache->setLocalIdsForGroup(groupSize, perThreadData.data(), *gfxCoreHelper.get());
+    localIdsCache->setLocalIdsForGroup(groupSize, perThreadData.data());
 
     EXPECT_EQ(groupSize, localIdsCache->cache[1].groupSize);
     EXPECT_NE(nullptr, localIdsCache->cache[1].localIdsData);
@@ -54,8 +50,7 @@ TEST_F(LocalIdsCacheTest, GivenEntryInCacheWhenGetLocalIdsForGroupThenEntryFromC
     localIdsCache->cache[0].localIdsSize = 512U;
     localIdsCache->cache[0].localIdsSizeAllocated = 512U;
     localIdsCache->cache[0].accessCounter = 1U;
-    auto gfxCoreHelper = NEO::GfxCoreHelper::create(NEO::defaultHwInfo->platform.eRenderCoreFamily);
-    localIdsCache->setLocalIdsForGroup(groupSize, perThreadData.data(), *gfxCoreHelper.get());
+    localIdsCache->setLocalIdsForGroup(groupSize, perThreadData.data());
     EXPECT_EQ(2U, localIdsCache->cache[0].accessCounter);
 }
 
@@ -68,8 +63,7 @@ TEST_F(LocalIdsCacheTest, GivenEntryWithBiggerBufferAllocatedWhenGetLocalIdsForG
     const auto localIdsData = localIdsCache->cache[0].localIdsData;
 
     groupSize = {2, 1, 1};
-    auto gfxCoreHelper = NEO::GfxCoreHelper::create(NEO::defaultHwInfo->platform.eRenderCoreFamily);
-    localIdsCache->setLocalIdsForGroup(groupSize, perThreadData.data(), *gfxCoreHelper.get());
+    localIdsCache->setLocalIdsForGroup(groupSize, perThreadData.data());
     EXPECT_EQ(1U, localIdsCache->cache[0].accessCounter);
     EXPECT_EQ(192U, localIdsCache->cache[0].localIdsSize);
     EXPECT_EQ(512U, localIdsCache->cache[0].localIdsSizeAllocated);