5. Known Issues and Workarounds
Known Issues and Workarounds in FPGA AI Suite Version 2025.1.1
There are no additional known issues in FPGA AI Suite Version 2025.1.1. The known issues from FPGA AI Suite Version 2025.1 still apply to Version 2025.1.1.
Known Issues and Workarounds in FPGA AI Suite Version 2025.1
FPGA AI Suite Version 2025.1 has the following known issues:
- The hostless design examples that use JTAG for host-FPGA communication may see a "claim_service: Path cannot be found" error.
The workaround is to reprogram the FPGA and lower the JTAG clock frequency to a lower frequency, such as 16 MHz, with a command similar to the following example command:
jtagconfig --setparam 1 JtagClock 16M
- The build example design flow might generate bitstreams with timing failures for the Agilex™ 5 hostless JTAG design examples. Inference performed on such bitstreams hangs.
When such faulty bitstreams are generated, the dla_adjust_pll.log log file found under the hw subfolder of the Quartus build directory contains the following error message:
Error (23098): One or more blocks are configured incorrectly and will not have the desired functionality
To work around this issue, relaunch the example design compilation with a different seed.
- The hostless DDR-free design example Quartus® Prime System Console script generates performance counter values that are displayed with a prefix that indicates that the value is a hexadecimal number. Despite the hexadecimal prefix, the value displayed is a decimal value.
Streaming Flow Known Issues
- Only single-output graphs are supported.
Multilane Known Issues
- The num_lanes parameter takes only the values 1, 2, or 4.
- Multilane exhibits low performance when used with DDR when a graph has residual connections.
OpenVINO Open Model Zoo Known Issues
The Open Model Zoo graph conversion tool (omz_converter) may encounter the following errors when converting certain TensorFlow models such as MobileNetV3:
ValueError: Invalid filepath extension for saving. Please add either a `.keras` extension for the native Keras format (recommended) or a `.h5` extension. Use `model.export(filepath)` if you want to export a SavedModel for use with TFLite/TFServing/etc.
For a workaround, refer to "Preparing a Model" in FPGA AI Suite Getting Started Guide .