Hi all,
We would like to know if floating point data type is supported in hexagon sdk.I referred the ECNS module in the examples and in the capi_v2_dummy_ecns.cpp i am able to see a fixed_point check and it supports only fixed point.
Please let us know whether we can do floating point operations using hexagon SDK.Our FFT and adaptive filters co-efficients are in floating point.
Regards,
Ram
Hi,
Floating point operations can be done with SDK but you need be aware that the interaction with framework will always be fixed point. CAPIv2 supports floating point calculations.