# Get started with the Qualcomm IM SDK
This section provides instructions on how to start developing your software using the Qualcomm IM SDK.
Tab Qualcomm Linux
Tab Ubuntu
- *class* tabincludedirectiv
- ## Qualcomm IM SDK workflow
The figure shows the SDK workflow from setting up the device to developing your first application. For more information, click each component.
**Figure : Qualcomm IM SDK workflow**
## Prerequisites
- Ensure that your development kit has the latest Qualcomm Linux software.
[Verify the Qualcomm Linux version](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-251/set_up_the_device.html#verify-the-qualcomm-linux-version) and if it isn't the latest software [update the software](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-251/upgrade-rb3gen2-software.html).
Else, upgrade your development kit following the steps described in [Set up the device](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-251/set_up_the_device.html).
- Ensure that you download the matching SDKs to your host machine before starting AI/ML development.
The sample applications and AI procedures in this guide are compatible with [Supported component versions](https://docs.qualcomm.com/doc/80-70022-51/topic/introduction.html#section-xsk-zrn-tcc).
## Explore sample applications
You can explore sample applications using the Qdemo application, which is part of the software installation. You can run this GUI-based application directly on the target device without setting up a host.
For more details, see [Run sample applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-253/run-sample-apps-qdemo.html).
## Explore eSDK
Run the [sample
applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-50/example-applications.html)
that are part of the Qualcomm IM SDK package and [GStreamer command-line use cases](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-50/gstreamer-application-use-cases.html)
to understand the eSDK capabilities.
## Download and install eSDK
Note
For the Qualcomm Dragonwing^TM^ IQ-8275, a downloadable eSDK isn't available. However, if you are a registered user with access to the Dragonwing IQ-8275 software,
you have the option to [create a custom eSDK](https://docs.qualcomm.com/doc/80-70022-51/topic/build-the-platfrom-esdk.html#build-the-platfrom-esdk).
After upgrading the development kit to the latest Qualcomm Linux software, download the corresponding eSDK.
This release supports separate eSDKs for the Qualcomm Dragonwing^TM^ RB3 Gen 2 Vision Development Kit, the Qualcomm Dragonwing^TM^ RB3 Gen 2 Core Development Kit, and Qualcomm Dragonwing^TM^ IQ-9075 Evaluation Kit (EVK).
You can download the eSDK that corresponds to your specific development kit.
**Prerequisites**
- An Ubuntu 22.04 host machine with at least 100 GB of free space. To
set up the virtual machine running Ubuntu 22.04 OS on a Microsoft
Windows or an Apple^®^ Mac^®^, see [Set Up a Virtual Machine for Qualcomm Linux](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-41/vm-landing-page.html).
- `sudo` permission to run the commands.
- Ensure to configure your Git username and email by running the following commands:
git config --global user.name
Copy to clipboard
git config --global user.email
Copy to clipboard
- Ensure that all the [system requirements](https://docs.yoctoproject.org/scarthgap/ref-manual/system-requirements.html) are installed on the host.
**Procedure**
1. Download the eSDK from [Qualcomm public
archive](https://artifacts.codelinaro.org/ui/native/qli-ci/flashable-binaries/qimpsdk/).
1. Make directory:
mkdir
Copy to clipboard
2. Change directory:
cd
Copy to clipboard
3. Download the zipped file:
- For Dragonwing RB3 Gen 2 Vision Development Kit:
>
>
> - Ubuntu x86 architecture-based host machines:
>
>
>
> >
> >
> > wget https://artifacts.codelinaro.org/artifactory/qli-ci/flashable-binaries/qimpsdk/qcs6490-rb3gen2-vision-kit/x86-qcom-6.6.97-QLI.1.6-Ver.1.2.1_qim-product-sdk-esdk-2.1.1.zip
> > Copy to clipboard
> - Arm^®^ architecture-based host machines:
>
>
> wget https://artifacts.codelinaro.org/artifactory/qli-ci/flashable-binaries/qimpsdk/qcs6490-rb3gen2-vision-kit/arm-qcom-6.6.97-QLI.1.6-Ver.1.2.1_qim-product-sdk-esdk-2.1.1.zip
> Copy to clipboard
- For Dragonwing IQ-9075 EVK:
>
>
> - Ubuntu x86 architecture-based host machines:
>
>
>
> >
> >
> > wget https://artifacts.codelinaro.org/artifactory/qli-ci/flashable-binaries/qimpsdk/qcs9075-iq-9075-evk/x86-qcom-6.6.97-QLI.1.6-Ver.1.2.1_qim-product-sdk-esdk-2.1.1.zip
> > Copy to clipboard
> - Arm architecture-based host machines:
>
>
>
> >
> >
> > wget https://artifacts.codelinaro.org/artifactory/qli-ci/flashable-binaries/qimpsdk/qcs9075-iq-9075-evk/arm-qcom-6.6.97-QLI.1.6-Ver.1.2.1_qim-product-sdk-esdk-2.1.1.zip
> > Copy to clipboard
4. Unzip the Qualcomm IM SDK to a directory of your choice:
unzip x86-qcom-6.6.97-QLI.1.6-Ver.1.2.1_qim-product-sdk-esdk-2.1.1.zip
Copy to clipboard
5. Ensure that the eSDK installer is at `/target/qcs6490-rb3gen2-vision-kit/sdk/`.
The following screenshot is provided as an example; the version numbers may differ from those in your installation.
6. If you don't have the required write permissions to install the
eSDK in the directory, the installer alerts you and then
terminates the installation. Ensure that you set up the required
permissions in the directory and rerun the installer.
umask a+rx
Copy to clipboard
2. To install the eSDK, run the installer script from `/target/qcs6490-rb3gen2-vision-kit/sdk/`:
sh ./qcom-wayland-x86_64-qcom-multimedia-image-armv8-2a--toolchain-ext-1.6-ver.1.2.1.sh
Copy to clipboard
For example:
sh ./qcom-wayland-x86\_64-qcom-multimedia-image-armv8-2a-qcs6490-rb3gen2-vision-kit-toolchain-ext-1.6-ver.1.2.1.sh
3. To install the eSDK in your host machine, follow the instructions on the console:
- On the console, you will be prompted to specify the directory path for the eSDK installation. If you accept the default path, the eSDK will be installed in the user directory, as shown in the following screenshot:
- Alternatively, if you specify a custom path, such as `/local/mnt/workspace/Platform_eSDK_plus_QIM`, the eSDK will be installed in that directory.
Note
If packages such as diffstat, gawk, gcc, and g++ are missing during the eSDK installation, install all the dependencies mentioned in
[Build with GitHub for unregistered users](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-254/github_workflow_unregistered_users.html).
4. To verify if the installation is successful, check if the Qualcomm IM SDK layers are available at `/layers`.
Note
If you want to add new software or customize the SDK with additional libraries and components, see [Create a custom eSDK for application development](https://docs.qualcomm.com/doc/80-70022-51/topic/build-the-platfrom-esdk.html#build-the-platfrom-esdk).
5. Set up the `ESDK_ROOT` variable.
export ESDK_ROOT=
Copy to clipboard
For example:
export ESDK\_ROOT=/local/mnt/workspace/Platform\_eSDK\_plus\_QIM
You have successfully installed the Qualcomm IM SDK. Now, [create a sample application with the Qualcomm IM SDK](https://docs.qualcomm.com/doc/80-70022-51/topic/content-develop-your-first-application.html#content-develop-your-first-application).
## Next steps
- [Qualcomm IM SDK overview](https://docs.qualcomm.com/doc/80-70022-51/topic/introduction.html#introduction)
- [Build multimedia applications](https://docs.qualcomm.com/doc/80-70022-51/topic/application-development.html#application-development)
- [Run sample applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-253/run-sample-apps-qdemo.html)
- [Set up virtual machine](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-41/vm-landing-page.html)
- [Build custom eSDK](https://docs.qualcomm.com/doc/80-70022-51/topic/build-the-platfrom-esdk.html#build-the-platfrom-esdk)
- [Qualcomm public archive](https://artifacts.codelinaro.org/ui/native/qli-ci/flashable-binaries/qimpsdk/)
- [System requirements](https://docs.yoctoproject.org/scarthgap/ref-manual/system-requirements.html)
- [Build with GitHub for unregistered users](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-254/github_workflow_unregistered_users.html)
- [Create a sample application](https://docs.qualcomm.com/doc/80-70022-51/topic/content-develop-your-first-application.html#content-develop-your-first-application)
- *class* tabincludedirectiv
- ## Qualcomm IM SDK workflow
The figure shows the SDK workflow from setting up the device to developing your first application.
This SDK workflow is currently supported only on the [Ubuntu Server](https://ubuntu.com/server) edition. It is not available on the [Ubuntu Desktop](https://ubuntu.com/desktop) edition.
The Ubuntu Server edition operates without a graphical user interface; all interactions are performed through the command-line interface, typically accessed through SSH.
The sample applications included in the SDK use Wayland to render and display content.
**Figure : Qualcomm IM SDK workflow**
## Prerequisites
- Ensure that your development kit has the latest version of the compatible [Ubuntu Server image](https://people.canonical.com/~platform/images/qualcomm-iot/ubuntu-24.04/ubuntu-24.04-x07/ubuntu-server-24.04/).
Else, upgrade your development kit following the steps described in [Flash the software images to the device](https://docs.qualcomm.com/bundle/publicresource/topics/80-90441-1/Integrate_and_flash_software_2.html).
- Ensure that you download the matching SDKs to your host machine before starting AI/ML development.
The sample applications and AI procedures in this guide are compatible with [Supported component versions](https://docs.qualcomm.com/doc/80-70022-51/topic/introduction.html#supported-component-versions-ubuntu).
## Set up the environment
1. Reboot the device and log in to the serial console by following the steps provided in [Verify reboot and sign in to the RB3 Gen 2 Ubuntu console](https://docs.qualcomm.com/bundle/publicresource/topics/80-90441-1/Use_Ubuntu_on_RB3_Gen2_3.html#verify-reboot-and-sign-in-to-the-rb3-gen-2-ubuntu-console).
When connected, the device is accessible through its configured IP address.
2. Connect to the device through the SSH shell:
ssh ubuntu@
Copy to clipboard
Note
Enter `ubuntu` as the SSH password when you boot the device for the first time after flashing.
3. Run the following commands to install GStreamer tools, the Qualcomm IM SDK, and sample applications in both native (C/C++) and Python:
sudo su
Copy to clipboard
apt-add-repository ppa:ubuntu-qcom-iot/qcom-ppa && apt update && apt install weston-autostart && apt install gstreamer1.0-tools && apt install tensorflow-lite-qcom-apps && apt install gstreamer1.0-qcom-sample-apps && apt install gstreamer1.0-qcom-python-examples && apt install qnn-tools && apt install snpe-tools
Copy to clipboard
The Qualcomm IM SDK and sample application installation is now complete.
## Explore sample applications
The Qualcomm IM SDK sample applications show how to use the SDKs and develop end-to-end edge analytics use cases. For more information about the sample applications for Ubuntu,
see [Sample applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-50/example-applications.html)
## Next steps
- [Explore and run sample applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70022-50/example-applications.html)
- [Ubuntu server](https://ubuntu.com/server)
- [Ubuntu desktop](https://ubuntu.com/desktop)
Last Published: Jan 04, 2026
[Previous Topic
Qualcomm IM SDK overview](https://docs.qualcomm.com/bundle/publicresource/80-70022-51/topics/introduction.md) [Next Topic
Build multimedia applications with the Qualcomm IM SDK](https://docs.qualcomm.com/bundle/publicresource/80-70022-51/topics/application-development.md)