diff --git a/benchmark/README.md b/benchmark/README.md
index 29c7155a0f33ecf29a04d7c276f4531ae64e07a6..367013f0457f9bbb9ae1335ea63dce181316d444 100644
--- a/benchmark/README.md
+++ b/benchmark/README.md
@@ -61,7 +61,7 @@ All platforms use cuDNN-v5.1. We see that caffe is slower in this experiment, be
 
 **Notation**
 
-All the experiments in caffe use `caffe time` to execute, which does not include the time of parameter updating. While PaddlePaddle and TensorFlow contains this time. But, compared with the total time, the time of parameter updating is relatively little on single machine.
+All the single-GPU experiments in caffe use `caffe time` to calculate elapsed time, which does not include parameter updating time. However, both PaddlePaddle and TensorFlow experiments contain the parameter updating time. As compared with the total time, this part is relatively little on single machine, we can ignore it.
 
 In Tensorflow, they implement algorithm searching method instead of using the algorithm searching interface in cuDNN.
 
@@ -102,7 +102,7 @@ We use lstm network for text classfication to test benchmark.
 
 ### Dataset
 -  [IMDB](http://www.iro.umontreal.ca/~lisa/deep/data/imdb.pkl)
-- Sequence legth is 100. In fact, PaddlePaddle supports training with variable-length sequence, but TensorFlow needs to pad, we also pad sequence length to 100 in PaddlePaddle in order to compare.
+- Sequence length is 100. In fact, PaddlePaddle supports training with variable-length sequence, but TensorFlow needs to pad. Thus, we also pad sequence length to 100 in PaddlePaddle in order to compare.
 - Dictionary size=30000 
 - Peephole connection is used in `lstmemory` by default in PaddlePaddle. It is also configured in TensorFlow.
 
diff --git a/benchmark/figs/alexnet-4gpu.png b/benchmark/figs/alexnet-4gpu.png
index 864c11313d7eacc90910a902c6a3a39c2bef77cc..1f5dddb94fabc81f40d7048aa6469dfc7e319288 100644
Binary files a/benchmark/figs/alexnet-4gpu.png and b/benchmark/figs/alexnet-4gpu.png differ
diff --git a/benchmark/figs/googlenet-4gpu.png b/benchmark/figs/googlenet-4gpu.png
index 098ed35bf7763d34993357501f16b6d859c8733f..5a7d87bf33af88feb10d1c19170fc90fc6f2de1d 100644
Binary files a/benchmark/figs/googlenet-4gpu.png and b/benchmark/figs/googlenet-4gpu.png differ