diff --git a/include/test.hpp b/include/test.hpp
index 7667032..bbee2d5 100644
--- a/include/test.hpp
+++ b/include/test.hpp
@@ -1,14 +1,14 @@
 #ifndef _TEST_KERNELS_H_
 #define _TEST_KERNELS_H_
 
-#include <stdint.h>
+#include <iostream>
 
 namespace base16384 {
 class test {
  public:
   // base16384_test_kernels_basic is a demo calculation that implements
   // mod, bit, plus and mul calculations.
-  SYCL_EXTERNAL static uint8_t kernels_basic(uint8_t in);
+  SYCL_EXTERNAL static std::uint8_t kernels_basic(uint8_t in);
 };
 }  // namespace base16384
 
diff --git a/libs/test_kernels.cpp b/libs/test_kernels.cpp
index 94d9a73..f563c61 100644
--- a/libs/test_kernels.cpp
+++ b/libs/test_kernels.cpp
@@ -1,10 +1,9 @@
-#include <stdint.h>
-
+#include <iostream>
 #include <sycl/sycl.hpp>
 
 #include "test.hpp"
 
-SYCL_EXTERNAL uint8_t base16384::test::kernels_basic(uint8_t in) {
+SYCL_EXTERNAL std::uint8_t base16384::test::kernels_basic(uint8_t in) {
   in *= in;
   in %= 251;
   in ^= in >> 2;
diff --git a/tests/basic.cpp b/tests/basic.cpp
index 7b2a4ef..2fe9a49 100644
--- a/tests/basic.cpp
+++ b/tests/basic.cpp
@@ -100,6 +100,17 @@ int main() {
   for (int i = 0; i < std::min(N, 64); i++) std::cout << " " << static_cast<int>(data[i]);
   std::cout << "..." << std::endl;
 
+  // Verify GPU basic parallel result
+  for (int i = 0; i < N; i++) {
+    if (data[i] != cpu_data[i]) {
+      std::cerr << "GPU 基本并行结果验证失败：位置 " << i << " 期望值 "
+                << static_cast<int>(cpu_data[i]) << " 实际值 " << static_cast<int>(data[i])
+                << std::endl;
+      sycl::free(data, q);
+      return -1;
+    }
+  }
+
   std::copy(initial_data.cbegin(), initial_data.cend(), data);
 
   start_time = std::chrono::high_resolution_clock::now();
@@ -122,6 +133,17 @@ int main() {
   for (int i = 0; i < std::min(N, 64); i++) std::cout << " " << static_cast<int>(data[i]);
   std::cout << "..." << std::endl;
 
+  // Verify GPU advanced parallel result
+  for (int i = 0; i < N; i++) {
+    if (data[i] != cpu_data[i]) {
+      std::cerr << "GPU 高级并行结果验证失败：位置 " << i << " 期望值 "
+                << static_cast<int>(cpu_data[i]) << " 实际值 " << static_cast<int>(data[i])
+                << std::endl;
+      sycl::free(data, q);
+      return -1;
+    }
+  }
+
   sycl::free(data, q);
 
   return 0;