README.md in tensor_stream-0.7.0 vs README.md in tensor_stream-0.8.0

@@ -322,9 +322,44 @@
 b = tf.constant(2.0)
 result = a + b
 File.write("model.pbtext", result.graph.as_graph_def)
 ```
 
+## Performance notes
+
+Comparative performance with respect to other ruby libraries have not yet been performed. However it is
+notable that TruffleRuby and ruby-2.6.0-preview2 with the --jit flag performs considerably better with respect
+to previous versions of ruby(< 2.6)
+
+Benchmarks running samples/linear_regression.rb on an Intel(R) Core(TM) i5-6200U CPU @ 2.30GHz
+
+ruby 2.4
+
+```
+$ ruby -v
+ruby 2.4.0p0 (2016-12-24 revision 57164) [x86_64-linux]
+$ ruby samples/linear_regression.rb
+495 seconds 1000 epochs
+```
+
+ruby 2.6.0-preview2
+
+```
+$ ruby -v
+ruby 2.6.0preview2 (2018-05-31 trunk 63539) [x86_64-linux]
+$ ruby --jit samples/linear_regression.rb
+394 seconds 10000 epochs
+```
+
+truffleruby
+```
+$ ruby -v
+truffleruby 1.0.0-rc5, like ruby 2.4.4, GraalVM CE Native [x86_64-linux]
+219 seconds 10000 epochs
+```
+
+For training large networks that works on images, the opencl evaluator is the only way to go.
+
 ## Roadmap
 
 - Docs
 - Complete low-level op support
 - SciRuby evaluator