diff --git a/paddle/fluid/inference/api/analysis_predictor.cc b/paddle/fluid/inference/api/analysis_predictor.cc
index 6319d99ab81100bdb36240bdb13112dad594f705..f3fbb1d4d3442a417f743351f8b2f4dcadbb4d54 100644
--- a/paddle/fluid/inference/api/analysis_predictor.cc
+++ b/paddle/fluid/inference/api/analysis_predictor.cc
@@ -505,8 +505,6 @@ std::unique_ptr<PaddlePredictor> CreatePaddlePredictor<
       std::string flag = "--fraction_of_gpu_memory_to_use=" +
                          std::to_string(fraction_of_gpu_memory);
       flags.push_back(flag);
-      // use auto growth strategy here.
-      flags.push_back("--allocator_strategy=auto_growth");
       flags.push_back("--cudnn_deterministic=True");
       VLOG(3) << "set flag: " << flag;
       framework::InitGflags(flags);