diff --git a/vllm/engine/llm_engine.py b/vllm/engine/llm_engine.py
index f8b9c48bc9589..a85b746297315 100644
--- a/vllm/engine/llm_engine.py
+++ b/vllm/engine/llm_engine.py
@@ -854,9 +854,6 @@ def _process_model_outputs(
             request_outputs.append(request_output)
         return request_outputs
 
-    def finish_measurements(self):
-        self.model_executor.finish_measurements()
-
     def step(self) -> List[Union[RequestOutput, EmbeddingRequestOutput]]:
         """Performs one decoding iteration and returns newly generated results.
 
diff --git a/vllm/entrypoints/llm.py b/vllm/entrypoints/llm.py
index fc9f118ff14b2..62309ed345b1d 100644
--- a/vllm/entrypoints/llm.py
+++ b/vllm/entrypoints/llm.py
@@ -173,9 +173,6 @@ def set_tokenizer(
             self.llm_engine.tokenizer.tokenizer = get_cached_tokenizer(
                 tokenizer)
 
-    def finish_measurements(self):
-        self.llm_engine.finish_measurements()
-
     @overload  # LEGACY: single (prompt + optional token ids)
     def generate(
         self,
diff --git a/vllm/executor/habana_executor.py b/vllm/executor/habana_executor.py
index 80f8037a2d043..6b97fb2e381af 100644
--- a/vllm/executor/habana_executor.py
+++ b/vllm/executor/habana_executor.py
@@ -90,9 +90,6 @@ def initialize_cache(self, num_gpu_blocks: int, num_cpu_blocks) -> None:
         msg = f"init_cache_engine took {cache_init_m.get_summary_string()}"
         logger.info(msg)
 
-    def finish_measurements(self):
-        self.driver_worker.finish_measurements()
-
     def execute_model(
             self,
             execute_model_req: ExecuteModelRequest) -> List[SamplerOutput]:
diff --git a/vllm/executor/ray_habana_executor.py b/vllm/executor/ray_habana_executor.py
index 17e3414a96b57..9e0a89cbeb8aa 100644
--- a/vllm/executor/ray_habana_executor.py
+++ b/vllm/executor/ray_habana_executor.py
@@ -237,9 +237,6 @@ def _driver_execute_model(
         return self.driver_worker.execute_method("execute_model",
                                                  execute_model_req)
 
-    def finish_measurements(self):
-        self._run_workers("finish_measurements")
-
     def execute_model(
             self,
             execute_model_req: ExecuteModelRequest) -> List[SamplerOutput]:
diff --git a/vllm/worker/habana_model_runner.py b/vllm/worker/habana_model_runner.py
index 72aba42ae8553..ea8b05f80ab3f 100644
--- a/vllm/worker/habana_model_runner.py
+++ b/vllm/worker/habana_model_runner.py
@@ -1503,4 +1503,5 @@ def shutdown_inc(self):
             print('inc shutdown')
 
     def __del__(self):
+        self.finish_measurements()
         self.shutdown_inc()
diff --git a/vllm/worker/habana_worker.py b/vllm/worker/habana_worker.py
index 87122c03d3c8f..6afa88a6238f5 100644
--- a/vllm/worker/habana_worker.py
+++ b/vllm/worker/habana_worker.py
@@ -223,9 +223,6 @@ def _warm_up_model(self) -> None:
         # the model initialization and profiling.
         set_random_seed(self.model_config.seed)
 
-    def finish_measurements(self):
-        self.model_runner.finish_measurements()
-
     @property
     def do_metadata_broadcast(self) -> bool:
         return self.parallel_config.tensor_parallel_size > 1