European Distributed Deep Learning (EDDL) library. A general-purpose library initially developed to cover deep learning needs in healthcare use cases within the DeepHealth project.
If we have a code (usually in tensor) that depends on a specific compilation to work (i.e hardware), we should encapsulate it with its corresponding preprocessing directive.
The reason behind this is that Tensor::isCPU/isGPU/isFPGA() relies on the Tensor->device value which can be set by the user and lead to errors hard to debug.
Example:
if (A->isCPU()) { // CPU
// Do stuff
}
#ifdef cGPU
else if (A->isGPU()){ // GPU
// Do stuff
}
#endif
#ifdef cFPGA
else if (A->isFPGA()) { // FPGA
// do stuff
}
#endif
Add preprocessor directives for FPGA
If we have a code (usually in tensor) that depends on a specific compilation to work (i.e hardware), we should encapsulate it with its corresponding preprocessing directive.
The reason behind this is that
Tensor::isCPU/isGPU/isFPGA()
relies on theTensor->device
value which can be set by the user and lead to errors hard to debug.Example: