[SINGA-41] Support single node single GPU training

• Type: New Feature
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Assignee: Unassigned
• Reporter: Haibo Chen
• Created: 27/Jul/15 08:43
• Updated: 29/Dec/15 09:24
• Resolved: 29/Dec/15 09:24

Description

For new feature: support single node single GPU training.

The blob class:
1) add a member function
Dtype* mutable_xpu_data(); // make a decision(GPU or CPU) at compile time
2) add a include file: #include <cuda_runtime.h> //support cuda code

The param class:
1) add three member function
float* mutable_xpu_data()
float* mutable_xpu_grad()
float* mutable_xpu_history()

The layer class:
1) we use xpu instead of cpu in some layers of subclass,eg.ConvolutionLayer、Dropout...

Makefile.GPU:
1) For CPU mode,set CUDA_DIR :=
step 1:
make
2) For GPU mode,set CUDA_DIR := your cuda installation path
step 1:
make gpu // rename layer file : layer.cc -> layer.cu
step 2:
make