Add warning for many core systems in Windows (#754)

* Add warning for many core systems in Windows

* Add warning for not busy cores in Windows

R-package/README.md

@@ -20,7 +20,7 @@ To force the usage of Rtools / MinGW, you can set `use_mingw` to `TRUE` in `R-pa
 
 For users who wants to install online with GPU or want to choose a specific compiler, please check the end of this document for installation using a helper package ([Laurae2/lgbdl](https://github.com/Laurae2/lgbdl/)).
 
-It is recommended to use *Visual Studio* for its better multi-threading efficency in Windows for many core systems. For very simple systems (dual core computers or worse), MinGW64 is recommended for maximum performance. If you do not know what to choose, it is recommended to use [Visual Studio](https://www.visualstudio.com/downloads/#build-tools-for-visual-studio-2017), the default compiler.
+**Warning for Windows users**: it is recommended to use *Visual Studio* for its better multi-threading efficency in Windows for many core systems. For very simple systems (dual core computers or worse), MinGW64 is recommended for maximum performance. If you do not know what to choose, it is recommended to use [Visual Studio](https://www.visualstudio.com/downloads/#build-tools-for-visual-studio-2017), the default compiler. **Do not try using MinGW in Windows on many core systems. It may result in 10x slower results than Visual Studio.**
 
 #### Mac OS X Preparation
 

docs/FAQ.md

@@ -80,6 +80,12 @@ Remember this is a free/open community support. We may not be available 24/7 to
 
 ---
 
+- **Question 8**: CPU are not kept busy (like 10% CPU usage only) in Windows when using LightGBM on very large datasets with many core systems.
+
+- **Solution 8**: Please use [use Visual Studio](https://github.com/Microsoft/LightGBM/issues/740) as it may be 10x faster than MinGW especially for very large trees.
+
+---
+
 ### R-package
 
 - **Question 1**: Any training command using LightGBM does not work after an error occurred during the training of a previous LightGBM model.