Find some fp16 bugs during decoding NMT system. We decide to roll back some codes to the last version. These codes need to be reviewed.

1. Fix the bugs in backward process.
2. Support the float16 class.
3. Fix bugs.
4. Clean the codes.
5. Remove the makefile.

Support fp16 data type for more operations and fix the minor errors.  (Don't use this! It's an incomplete version)

Support fp16 data type for more operations and fix the minor errors.  (Don't use this! It's an incomplete version)

Fix the bugs in FlushToMem functions.

1. Fix the bugs in SetDevice functions and FlushToMem functions.
2. Fix the minor errors.

1. Move the SetAscendingOrder functions from XTensor.*  to core/utilities/SetAscendingOder.*
2. Rename XElement.* files to XCall.*

1. Move the New/Initial/Del Tensor functions from XTensor.* to XElement.*
2. Move IsSameShaped functions from XTensor.* to core/shape/IsSameShaped.*
3. Move CheckData functions from XTensor.* to core/utilities/CheckData.*

1. Merge with Huchi branch (replace all requireLink with enableGrad that allows gradient computation for a tensor);
2. Update the global memory size (This may make the memory size a little bit larger than the old version).

2. Add MemManager class to manage the memory;
3. Use the new version XList;
4. Add a new interface to create Tensor;
5. Apply MemManager to FNNLM.

1. add macro to implement unary function 2. add sub and div function 3. merge code with the latest branch of xiaotong-working

1. Bug fixed; 2. Support abs, sign, convert data type(float <->int) ;3. Merge with xu and xiao branch