diff --git a/src/gpu-compute/global_memory_pipeline.cc b/src/gpu-compute/global_memory_pipeline.cc
index bcd93f8862..a2b24e4b47 100644
--- a/src/gpu-compute/global_memory_pipeline.cc
+++ b/src/gpu-compute/global_memory_pipeline.cc
@@ -130,6 +130,9 @@ GlobalMemPipeline::exec()
         DPRINTF(GPUMem, "CU%d: WF[%d][%d]: Completing global mem instr %s\n",
                 m->cu_id, m->simdId, m->wfSlotId, m->disassemble());
         m->completeAcc(m);
+        if (m->isFlat() && m->isLoad()) {
+            w->decLGKMInstsIssued();
+        }
         w->decVMemInstsIssued();
 
         if (m->isLoad() || m->isAtomicRet()) {
@@ -193,6 +196,10 @@ GlobalMemPipeline::exec()
                 mp->disassemble(), mp->seqNum());
         mp->initiateAcc(mp);
 
+        if (mp->isFlat() && mp->isStore()) {
+            mp->wavefront()->decLGKMInstsIssued();
+        }
+
         if (mp->isStore() && mp->isGlobalSeg()) {
             mp->wavefront()->decExpInstsIssued();
         }
diff --git a/src/gpu-compute/gpu_dyn_inst.cc b/src/gpu-compute/gpu_dyn_inst.cc
index 03ed689519..38e4ecf492 100644
--- a/src/gpu-compute/gpu_dyn_inst.cc
+++ b/src/gpu-compute/gpu_dyn_inst.cc
@@ -819,7 +819,6 @@ GPUDynInst::resolveFlatSegment(const VectorMask &mask)
     if (executedAs() == Enums::SC_GLOBAL) {
         // no transormation for global segment
         wavefront()->execUnitId =  wavefront()->flatGmUnitId;
-        wavefront()->decLGKMInstsIssued();
         if (isLoad()) {
             wavefront()->rdLmReqsInPipe--;
         } else if (isStore()) {