eIQ Toolkit does not work in Ubuntu 22.04

[deneriz]

I've downloaded eIQ Toolkit 1.16.0.106 Ubuntu Installer and installed it with:

chmod +x eiq-toolkit-v1.16.0.106-1_amd64_b250703.deb.bin
sudo ./eiq-toolkit-v1.16.0.106-1_amd64_b250703.deb.bin

If I launch the eIQ Toolkit portal I  get the following:

source /opt/nxp/eIQ_Toolkit_v1.16.0/bin/eiqenv.sh && /opt/nxp/eIQ_Toolkit_v1.16.0/eiq-portal
eIQ Portal version 2.16.0 
------------------------------------------------
/opt/nxp/eIQ_Toolkit_v1.16.0/resources/app.asar
------------------------------------------------
Launch -> /opt/nxp/eIQ_Toolkit_v1.16.0
is-elevated: false
09:36:34.254 › Launching Application
Display size is 1920x1080
[2025-08-01T07:36:34.582Z] ExtensionHostController running as pid 23867
[2025-08-01T07:36:35.018Z] InstalledExtensionsReady barrier opened
[2025-08-01T07:36:35.049Z] Socket.io logging server connected to socket COKliSleb6yc_PxXAAAB
(buffered) [2025-08-01T07:36:34.906Z] ExtensionHost running as pid 23976
(buffered) [2025-08-01T07:36:35.036Z] Connected to sqlite:///home/deneriz/.config/eiqportal-nodejs/state.sqlite
09:36:35.133 › [CONVERTER] 2025-08-01 09:36:35.129579: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:477] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered

09:36:35.144 › [CONVERTER] WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1754033795.144002   24023 cuda_dnn.cc:8310] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered

09:36:35.148 › [CONVERTER] E0000 00:00:1754033795.148069   24023 cuda_blas.cc:1418] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered

09:36:35.162 › [CONVERTER] 2025-08-01 09:36:35.161826: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.

09:36:35.238 › [TRAINER] 2025-08-01 09:36:35.237425: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:477] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered

09:36:35.253 › [TRAINER] WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1754033795.252816   24037 cuda_dnn.cc:8310] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered

09:36:35.257 › [TRAINER] E0000 00:00:1754033795.257420   24037 cuda_blas.cc:1418] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered

09:36:35.273 › [TRAINER] 2025-08-01 09:36:35.273235: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.

09:36:35.319 › [IMPORTER] 2025-08-01 09:36:35.318661: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:477] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered

09:36:35.335 › [IMPORTER] WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1754033795.335405   24038 cuda_dnn.cc:8310] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered

09:36:35.340 › [IMPORTER] E0000 00:00:1754033795.340386   24038 cuda_blas.cc:1418] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered

09:36:35.355 › [IMPORTER] 2025-08-01 09:36:35.354815: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.

[2025-08-01T07:36:35.435Z] Socket.io logging server connected to socket qQ74H5beOrLxNSpSAAAD
libva error: vaGetDriverNameByIndex() failed with unknown libva error, driver_name = (null)
libva error: vaGetDriverNameByIndex() failed with unknown libva error, driver_name = (null)
libva error: vaGetDriverNameByIndex() failed with unknown libva error, driver_name = (null)
libva error: vaGetDriverNameByIndex() failed with unknown libva error, driver_name = (null)
  0%|          | 0/1 [00:00<?, ?it/s]
[23867:0801/093636.004460:FATAL:gpu_data_manager_impl_private.cc(415)] GPU process isn't usable. Goodbye.

However, my GPU is CUDA usable:

nvidia-smi
Fri Aug  1 09:37:00 2025       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 565.57.01              Driver Version: 565.57.01      CUDA Version: 12.7     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 4060 ...    On  |   00000000:01:00.0  On |                  N/A |
| N/A   43C    P4             11W /   55W |    1054MiB /   8188MiB |     56%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
                                                                                        

I'm on Ubuntu 22.04:

lsb_release -a
No LSB modules are available.
Distributor ID:	Ubuntu
Description:	Ubuntu 22.04.5 LTS
Release:	22.04
Codename:	jammy

Where can I find the official installation instructions for eIQ Toolkit 1.16? Is there a way to launch the eIQ Toolkit without GPU? Why the eIQ Toolkit Portal not booting?

[danielchen]

Hi    deneriz

Can you try to use command line as below?

$ source /opt/nxp/eIQ_Toolkit_v1.14.0/bin/eiqenv.sh

$ eiq-converter -h

Regards

Daniel

[deneriz · Accepted solution]

It seems the eIQ Toolkit on Linux must be used through CLI. I've been able to use it in Ubuntu 22.04.5 LTS. As pointed out by @danielchen here, after downloading it you have to setup the environment with:

source /opt/nxp/eIQ_Toolkit_v1.16.0/bin/eiqenv.sh

And you will have access to the main converter tool:

eiq-converter --help
eiq-converter 0.0.0
usage: eiq-converter [options]
options:
    [-h] [--help] [plugin_name source_type dest_type]      Prints this help for a plugin or this help
    [-v] [--version]                             Prints version
    [-r] [--reserve]                             Prints Reserved Params List
    [-l] [--list]                                Lists installed plugins
    [-i] [--install] path_to_whl_file            Installs a plugin
    [-u] [--uninstall] plugin_name               Uninstalls a plugin
    [--source]  source_model [--dest] destination_model [conveter specific parameters]    Converts a model

You will find documentation about how to use the CLI at:

/opt/nxp/eIQ_Toolkit_v1.16.0/docs/eIQ_Toolkit_UG.pdf