diff --git a/level_zero/core/source/cmdlist/cmdlist_hw_base.inl b/level_zero/core/source/cmdlist/cmdlist_hw_base.inl
index c18098d723..1dab1577f8 100644
--- a/level_zero/core/source/cmdlist/cmdlist_hw_base.inl
+++ b/level_zero/core/source/cmdlist/cmdlist_hw_base.inl
@@ -95,9 +95,6 @@ ze_result_t CommandListCoreFamily<gfxCoreFamily>::appendLaunchKernelWithParams(z
         kernel->setGroupCount(pThreadGroupDimensions->groupCountX,
                               pThreadGroupDimensions->groupCountY,
                               pThreadGroupDimensions->groupCountZ);
-        kernel->patchWorkDim(pThreadGroupDimensions->groupCountX,
-                             pThreadGroupDimensions->groupCountY,
-                             pThreadGroupDimensions->groupCountZ);
     }
 
     if (isIndirect && pThreadGroupDimensions) {
diff --git a/level_zero/core/source/cmdlist/cmdlist_hw_xehp_and_later.inl b/level_zero/core/source/cmdlist/cmdlist_hw_xehp_and_later.inl
index e253e1c208..58d3b16b95 100644
--- a/level_zero/core/source/cmdlist/cmdlist_hw_xehp_and_later.inl
+++ b/level_zero/core/source/cmdlist/cmdlist_hw_xehp_and_later.inl
@@ -146,9 +146,6 @@ ze_result_t CommandListCoreFamily<gfxCoreFamily>::appendLaunchKernelWithParams(z
         kernel->setGroupCount(pThreadGroupDimensions->groupCountX,
                               pThreadGroupDimensions->groupCountY,
                               pThreadGroupDimensions->groupCountZ);
-        kernel->patchWorkDim(pThreadGroupDimensions->groupCountX,
-                             pThreadGroupDimensions->groupCountY,
-                             pThreadGroupDimensions->groupCountZ);
     }
     NEO::GraphicsAllocation *eventAlloc = nullptr;
     uint64_t eventAddress = 0;
diff --git a/level_zero/core/source/kernel/kernel.h b/level_zero/core/source/kernel/kernel.h
index 77f906f08b..f966bb7e43 100644
--- a/level_zero/core/source/kernel/kernel.h
+++ b/level_zero/core/source/kernel/kernel.h
@@ -116,8 +116,6 @@ struct Kernel : _ze_kernel_handle_t, virtual NEO::DispatchKernelEncoderI {
     virtual ze_result_t setGlobalOffsetExp(uint32_t offsetX, uint32_t offsetY, uint32_t offsetZ) = 0;
     virtual void patchGlobalOffset() = 0;
 
-    virtual void patchWorkDim(uint32_t groupCountX, uint32_t groupCountY, uint32_t groupCountZ) = 0;
-
     virtual ze_result_t suggestMaxCooperativeGroupCount(uint32_t *totalGroupCount, NEO::EngineGroupType engineGroupType,
                                                         bool isEngineInstanced) = 0;
     virtual ze_result_t setCacheConfig(ze_cache_config_flags_t flags) = 0;
diff --git a/level_zero/core/source/kernel/kernel_imp.cpp b/level_zero/core/source/kernel/kernel_imp.cpp
index a12d6b450e..c17005e65d 100644
--- a/level_zero/core/source/kernel/kernel_imp.cpp
+++ b/level_zero/core/source/kernel/kernel_imp.cpp
@@ -280,6 +280,18 @@ void KernelImp::setGroupCount(uint32_t groupCountX, uint32_t groupCountY, uint32
 
     uint32_t groupCount[3] = {groupCountX, groupCountY, groupCountZ};
     NEO::patchVecNonPointer(dst, desc.payloadMappings.dispatchTraits.numWorkGroups, groupCount);
+
+    uint32_t workDim = 1;
+    if (groupCountZ * groupSize[2] > 1) {
+        workDim = 3;
+    } else if (groupCountY * groupSize[1] > 1) {
+        workDim = 2;
+    }
+    auto workDimOffset = desc.payloadMappings.dispatchTraits.workDim;
+    if (NEO::isValidOffset(workDimOffset)) {
+        auto destinationBuffer = ArrayRef<uint8_t>(crossThreadData.get(), crossThreadDataSize);
+        NEO::patchNonPointer(destinationBuffer, desc.payloadMappings.dispatchTraits.workDim, workDim);
+    }
 }
 
 ze_result_t KernelImp::setGroupSize(uint32_t groupSizeX, uint32_t groupSizeY,
@@ -924,21 +936,6 @@ void KernelImp::patchGlobalOffset() {
     NEO::patchVecNonPointer(dst, desc.payloadMappings.dispatchTraits.globalWorkOffset, this->globalOffsets);
 }
 
-void KernelImp::patchWorkDim(uint32_t groupCountX, uint32_t groupCountY, uint32_t groupCountZ) {
-    const NEO::KernelDescriptor &kernelDescriptor = kernelImmData->getDescriptor();
-    auto dataOffset = kernelDescriptor.payloadMappings.dispatchTraits.workDim;
-    if (NEO::isValidOffset(dataOffset)) {
-        auto destinationBuffer = ArrayRef<uint8_t>(crossThreadData.get(), crossThreadDataSize);
-        uint32_t workDim = 1;
-        if (groupCountZ * groupSize[2] > 1) {
-            workDim = 3;
-        } else if (groupCountY * groupSize[1] > 1) {
-            workDim = 2;
-        }
-        NEO::patchNonPointer(destinationBuffer, kernelDescriptor.payloadMappings.dispatchTraits.workDim, workDim);
-    }
-}
-
 Kernel *Kernel::create(uint32_t productFamily, Module *module,
                        const ze_kernel_desc_t *desc, ze_result_t *res) {
     UNRECOVERABLE_IF(productFamily >= IGFX_MAX_PRODUCT);
diff --git a/level_zero/core/source/kernel/kernel_imp.h b/level_zero/core/source/kernel/kernel_imp.h
index 72868b5a18..c3beaf2b02 100644
--- a/level_zero/core/source/kernel/kernel_imp.h
+++ b/level_zero/core/source/kernel/kernel_imp.h
@@ -127,8 +127,6 @@ struct KernelImp : Kernel {
     ze_result_t setGlobalOffsetExp(uint32_t offsetX, uint32_t offsetY, uint32_t offsetZ) override;
     void patchGlobalOffset() override;
 
-    void patchWorkDim(uint32_t groupCountX, uint32_t groupCountY, uint32_t groupCountZ) override;
-
     ze_result_t setCacheConfig(ze_cache_config_flags_t flags) override;
     bool usesRayTracing() {
         return kernelImmData->getDescriptor().hasRTCalls();
diff --git a/level_zero/core/test/unit_tests/sources/kernel/test_kernel.cpp b/level_zero/core/test/unit_tests/sources/kernel/test_kernel.cpp
index faa1b5f8a1..8077fd6f76 100644
--- a/level_zero/core/test/unit_tests/sources/kernel/test_kernel.cpp
+++ b/level_zero/core/test/unit_tests/sources/kernel/test_kernel.cpp
@@ -1929,8 +1929,6 @@ HWTEST_F(KernelWorkDimTests, givenGroupCountsWhenPatchingWorkDimThenCrossThreadD
     createKernel(kernel.get());
     kernel->setCrossThreadData(sizeof(uint32_t));
 
-    kernel->patchWorkDim(1, 1, 1);
-
     mockKernelImmData->mockKernelDescriptor->payloadMappings.dispatchTraits.workDim = 0x0u;
 
     auto destinationBuffer = ArrayRef<const uint8_t>(kernel->getCrossThreadData(), kernel->getCrossThreadDataSize());
@@ -1950,8 +1948,7 @@ HWTEST_F(KernelWorkDimTests, givenGroupCountsWhenPatchingWorkDimThenCrossThreadD
     for (auto &[groupSizeX, groupSizeY, groupSizeZ, groupCountX, groupCountY, groupCountZ, expectedWorkDim] : sizesCountsWorkDim) {
         ze_result_t res = kernel->setGroupSize(groupSizeX, groupSizeY, groupSizeZ);
         EXPECT_EQ(res, ZE_RESULT_SUCCESS);
-
-        kernel->patchWorkDim(groupCountX, groupCountY, groupCountZ);
+        kernel->setGroupCount(groupCountX, groupCountY, groupCountZ);
         EXPECT_EQ(*workDimInCrossThreadDataPtr, expectedWorkDim);
     }
 }