diff --git a/ci/jenkins/docker-images.ini b/ci/jenkins/docker-images.ini
index a1c1ed5238da..a93e272ce88b 100644
--- a/ci/jenkins/docker-images.ini
+++ b/ci/jenkins/docker-images.ini
@@ -20,7 +20,7 @@
 ci_arm: tlcpack/ci-arm:20230314-060145-ccc0b9162
 ci_cortexm: tlcpackstaging/ci_cortexm:20230124-233207-fd3f8035c
 ci_cpu: tlcpack/ci-cpu:20230308-070109-9d732d0fa
-ci_gpu: tlcpack/ci-gpu:20230308-070109-9d732d0fa
+ci_gpu: tlcpack/ci-gpu:20230318-060139-2ff41c615
 ci_hexagon: tlcpack/ci_hexagon:20230127-185848-95fa22308
 ci_i386: tlcpack/ci-i386:20221013-060115-61c9742ea
 ci_lint: tlcpack/ci-lint:20221013-060115-61c9742ea
diff --git a/src/target/target_kind.cc b/src/target/target_kind.cc
index a87bb92c483b..d1b2c10edf01 100644
--- a/src/target/target_kind.cc
+++ b/src/target/target_kind.cc
@@ -161,8 +161,8 @@ TargetJSON UpdateCUDAAttrs(TargetJSON target) {
     // Use the compute version of the first CUDA GPU instead
     TVMRetValue version;
     if (!DetectDeviceFlag({kDLCUDA, 0}, runtime::kComputeVersion, &version)) {
-      LOG(WARNING) << "Unable to detect CUDA version, default to \"-arch=sm_20\" instead";
-      archInt = 20;
+      LOG(WARNING) << "Unable to detect CUDA version, default to \"-arch=sm_50\" instead";
+      archInt = 50;
     } else {
       archInt = std::stod(version.operator std::string()) * 10 + 0.1;
     }
@@ -189,8 +189,8 @@ TargetJSON UpdateNVPTXAttrs(TargetJSON target) {
     // Use the compute version of the first CUDA GPU instead
     TVMRetValue version;
     if (!DetectDeviceFlag({kDLCUDA, 0}, runtime::kComputeVersion, &version)) {
-      LOG(WARNING) << "Unable to detect CUDA version, default to \"-mcpu=sm_20\" instead";
-      arch = 20;
+      LOG(WARNING) << "Unable to detect CUDA version, default to \"-mcpu=sm_50\" instead";
+      arch = 50;
     } else {
       arch = std::stod(version.operator std::string()) * 10 + 0.1;
     }
diff --git a/tests/python/frontend/tensorflow/test_forward.py b/tests/python/frontend/tensorflow/test_forward.py
index 1ca0f3faef47..703df799423f 100644
--- a/tests/python/frontend/tensorflow/test_forward.py
+++ b/tests/python/frontend/tensorflow/test_forward.py
@@ -249,6 +249,8 @@ def compare_tf_with_tvm(
     targets=None,
     ignore_in_shape=False,
     convert_config=None,
+    atol=1e-5,
+    rtol=1e-5,
 ):
     """Generic function to generate and compare tensorflow and TVM output"""
 
@@ -303,7 +305,7 @@ def name_without_num(name):
             for i, tf_out in enumerate(tf_output):
                 if not isinstance(tf_out, np.ndarray):
                     assert len(tvm_output[i].shape) == 0  # pylint: disable=len-as-condition
-                tvm.testing.assert_allclose(tf_out, tvm_output[i], atol=1e-5, rtol=1e-5)
+                tvm.testing.assert_allclose(tf_out, tvm_output[i], atol=atol, rtol=rtol)
 
         sess.close()
 
@@ -3401,6 +3403,8 @@ def _test_forward_crop_and_resize(
     extrapolation_value=0.0,
     method="bilinear",
     dtype="float32",
+    atol=1e-4,
+    rtol=1e-4,
 ):
     image = np.random.uniform(0, 10, size=img_shape).astype(dtype)
     tf.reset_default_graph()
@@ -3415,7 +3419,7 @@ def _test_forward_crop_and_resize(
             extrapolation_value=extrapolation_value,
             name="crop_and_resize",
         )
-        compare_tf_with_tvm([image], ["in_data:0"], "crop_and_resize:0")
+        compare_tf_with_tvm([image], ["in_data:0"], "crop_and_resize:0", atol=atol, rtol=rtol)
 
 
 def test_forward_crop_and_resize():
@@ -3444,6 +3448,8 @@ def test_forward_crop_and_resize():
         box_idx=[1, 0, 2, 3],
         crop_size=[24, 24],
         extrapolation_value=0.3,
+        atol=1e-3,
+        rtol=1e-3,
     )
     _test_forward_crop_and_resize(
         img_shape=[20, 229, 229, 3],
@@ -3452,6 +3458,8 @@ def test_forward_crop_and_resize():
         crop_size=[58, 58],
         extrapolation_value=0.2,
         method="nearest",
+        atol=1e-3,
+        rtol=1e-3,
     )
 
 
diff --git a/tests/scripts/request_hook/request_hook.py b/tests/scripts/request_hook/request_hook.py
index e1b6cc7d594d..323443f12940 100644
--- a/tests/scripts/request_hook/request_hook.py
+++ b/tests/scripts/request_hook/request_hook.py
@@ -109,6 +109,7 @@
     "https://github.com/dmlc/web-data/raw/main/gluoncv/detection/street_small.jpg": f"{BASE}/2022-10-05/gluon-small-stree.jpg",
     "https://github.com/dmlc/web-data/raw/main/tensorflow/models/Custom/placeholder.pb": f"{BASE}/dmlc/web-data/raw/main/tensorflow/models/Custom/placeholder.pb",
     "https://github.com/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/classify_image_graph_def-with_shapes.pb": f"{BASE}/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/classify_image_graph_def-with_shapes.pb",
+    "https://github.com/dmlc/web-data/raw/main/tensorflow/models/ResnetV2/resnet-20180601_resnet_v2_imagenet-shapes.pb": f"{BASE}/dmlc/web-data/raw/main/tensorflow/models/ResnetV2/resnet-20180601_resnet_v2_imagenet-shapes.pb",
     "https://github.com/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/elephant-299.jpg": f"{BASE}/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/elephant-299.jpg",
     "https://github.com/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/imagenet_2012_challenge_label_map_proto.pbtxt": f"{BASE}/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/imagenet_2012_challenge_label_map_proto.pbtxt",
     "https://github.com/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/imagenet_synset_to_human_label_map.txt": f"{BASE}/dmlc/web-data/raw/main/tensorflow/models/InceptionV1/imagenet_synset_to_human_label_map.txt",