Forum Discussion

RubenPadial's avatar
RubenPadial
Icon for Contributor rankContributor
2 years ago

Intel FPGA AI suite input data type

Hello, I'm using Intel FPGA AI 2023.2 on ubuntu 20.04 host computer and trying to infer a custom CNN in a Intel Arria 10 SoC FPGA. I have followed Intel FPGA AI Suite SoC Design Example Guide and...
Acceleration
RubenPadial's avatar
RubenPadial
Icon for Contributor rankContributor
2 years ago

Hello@JohnT_Intel,

I prefer to manage the issues in different threads as they seem to have different root causes.

RubenPadial's avatar
RubenPadial
Icon for Contributor rankContributor
2 years ago

Hello @JohnT_Intel ,

Is there any news? The same issue is pressent when importing from ONNX model.

ONNX is imported to IR model and then Input layer is detected as U8 layer in the compiled graph. It should be FP32 instead.

  • RubenPadial's avatar
    RubenPadial
    Icon for Contributor rankContributor
    2 years ago

    Hello,

    It seems fixed in Intel FPGA AI Suite 2023.3.

    In Intel FPGA AI Suite 2023.2 input datatype is not modified by using "--bin-data" dla_compiler flag.