diff --git a/docs/source/quick_start/md/quick_start.md b/docs/source/quick_start/md/quick_start.md
index de50b83037db464787b45e86b0344f536a07742e..7df4f48a997e35c001ae462b1f9d3052434eb443 100644
--- a/docs/source/quick_start/md/quick_start.md
+++ b/docs/source/quick_start/md/quick_start.md
@@ -49,7 +49,7 @@ Currently our PGL is developed based on static computational mode of paddle (we
 import paddle.fluid as fluid
 
 use_cuda = False  
-place = fluid.GPUPlace(0) if use_cuda else fluid.CPUPlace()
+place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
 
 # use GraphWrapper as a container for graph data to construct a graph neural network
 gw = pgl.graph_wrapper.GraphWrapper(name='graph',