[GPU] Do not preallocate memory if desired size reaches the maximum s…

…upported buffer size

src/plugins/intel_gpu/src/runtime/shape_predictor.cpp

@@ -53,6 +53,9 @@ bool ShapePredictor::can_preallocate(size_t desired_buffer_size) {
     const auto memory_threshold = 0.90f;
     auto device_mem_usage = _engine->get_used_device_memory(cldnn::allocation_type::usm_device);
 
+    if (desired_buffer_size > _engine->get_device_info().max_alloc_mem_size)
+        return false;
+
     return device_mem_usage + desired_buffer_size < _engine->get_device_info().max_global_mem_size * memory_threshold;
 }
 

src/plugins/intel_gpu/tests/unit/module_tests/shape_predictor_test.cpp

@@ -117,3 +117,23 @@ INSTANTIATE_TEST_SUITE_P(smoke, shape_predictor_tests_b_fs_yx_fsv16,
         {{{1,1}, {1,128}, {1,256}}, {}, 1.1f, true},
         {{{1,3,128}, {1,3,112}, {1,3,418}, {1,3,512}}, {}, 1.1f, true},
     }));
+
+TEST(shape_predictor_tests, check_max_buffer_size) {
+    auto& engine = get_test_engine();
+
+    const auto& buffers_preallocation_ratio = 1.1;
+    ShapePredictor sp(&engine, buffers_preallocation_ratio);
+
+    const auto max_alloc_mem_size = engine.get_device_info().max_alloc_mem_size;
+    auto layout = cldnn::layout({static_cast<int64_t>(max_alloc_mem_size)}, ov::element::u8, format::bfyx);
+
+    std::pair<bool, ov::Shape> result;
+
+    // Perform 3 iteration to trigger shape preallocation
+    result = sp.predict_preallocation_shape("dummy_name", layout, false);
+    result = sp.predict_preallocation_shape("dummy_name", layout, false);
+    result = sp.predict_preallocation_shape("dummy_name", layout, false);
+
+    const auto bytes_count = ov::shape_size(result.second);
+    ASSERT_FALSE(sp.can_preallocate(bytes_count));
+}