Did you know you can set up your Intel hardware to validate Edge AI use cases in about 5 minutes?

[stevenhoenisch]

Here's the fast path for setting up Intel hardware to validate a variety of Edge AI use cases with Edge AI Suites and, optionally, other components of Open Edge Platform.

First, make sure that your Intel hardware appears in the list of validated hardware and configurations and is running Ubuntu 24.04 LTS with the 6.14 kernel hardware enablement stack (HWE). The scripts in our Edge Developer Kit Reference Scripts repository were tested on Ubuntu 24.04.3 LTS, and we recommend using the HWE kernel that matches this tested version to avoid driver or packaging mismatches.

Then run the following script, reproduced below from the 5-minute quick start, to simplify and speed up the configuration for running edge AI workloads on various Intel platforms and GPUs:

sudo bash -c "$(wget -qLO - https://raw.githubusercontent.com/open-edge-platform/edge-developer-kit-reference-scripts/refs/heads/main/main_installer.sh)"

If you run into problems or want to read the step-by-step instructions, see Getting Started with Intel® Edge Developer Kits.

If all goes well, you should see an installation summary that looks something like the following. (If the summary doesn't appear, see the instructions for handling reboot.)

========================================================================
Running Installation Summary
==================== System Installation Summary ====================
Item                      | Value
------------------------ -+-----------------------------------------
Kernel Version            | 6.14.0-27-generic
HWE Stack                 | Installed
Ubuntu Version            | Ubuntu 24.04.3 LTS
NPU Status                | Detected
NPU Package               | intel-level-zero-npu
NPU Version               | 1.19.0.20250707-16111289554
intel-driver-compiler-npu | 1.19.0.20250707-16111289554
intel-fw-npu              | 1.19.0.20250707-16111289554
intel-level-zero-npu      | 1.19.0.20250707-16111289554
level-zero                | 1.22.4
GPU Type                  | Intel
GPU Count                 | 4 Intel graphics device(s) detected
GPU Driver                | i915 (loaded)
GPU Device 1              | 00:02.0 VGA compatible controller: Intel Corporation Arrow Lake-U [Intel Graphics] (rev 06)
GPU Device 2              | 03:00.0 VGA compatible controller: Intel Corporation Device e20b
GPU Device 3              | 08:00.0 VGA compatible controller: Intel Corporation Device e20b
GPU Device 4              | 80:14.5 Non-VGA unclassified device: Intel Corporation Device 7f2f (rev 10)
------------------------ -+-----------------------------------------
Intel Graphics Packages   |
------------------------ -+-----------------------------------------
i965-va-driver:amd64      | 2.4.1+dfsg1-1build2
intel-gsc                 | 0.9.5-0ubuntu1~24.04~ppa1
intel-media-va-driver-non-free:amd64 | 25.3.1-0ubuntu1~24.04~ppa1
intel-opencl-icd          | 25.27.34303.9-1~24.04~ppa1
libegl-mesa0:amd64        | 25.0.7-0ubuntu0.24.04.1
------------------------ -+-----------------------------------------
Platform Status           | [✓] Platform is configured
=====================================================================
Installation completed: 2025-08-11 10:11:54
Log file saved: /var/log/intel-platform-installer.log
========================================================================

⏭️ After the setup finishes, choose an Edge AI use case to quickly deploy a workflow or sample application, such as Metro AI Suite's smart parking sample app, and begin validating your Intel hardware. The use case you choose is, of course, up to you -- there are use cases suitable for AI beginners through experts -- but if you have a few more minutes, why not pull forward into smart parking? 🅿️ 🏎️