Hello,
I would like to know how to transfer data between cores (CPU, GPU, DSP (HVX), HTA, etc.) to configure a deep learning inference pipeline in the form of data load -> pre-processing -> SNPE inference -> post-processing.
I would also like to know how to divide processes by core.
If Qualcomm provides an API for this, please let me know the name of the API or related documentation.
Thnaks.
Dear developer,
There is tow way for transfering data from one backend to another.
1 General DDR
2 ION buffer also known as cDSP buffer.