`xpumd` fails with intel/xpumanager:v1.2.27 image, but succeeds with updated dependencies

The latest xpumanager container (which uses an 1.5 year old release) fails to initialize the zero API on my system. I'm running:
* One Raptor Lake iGPU (i9 13000H)
* One Arc Pro A40 dedicated
* Kernel version 6.12.18
* i915 driver

The container fails to initialize devices with:
```
[2025-06-14 01:26:05.324] [I] [1-1] XPUM: Init xpum library
[2025-06-14 01:26:05.324] [I] [1-1] XPU Manager: 1.2.27.20240130
[2025-06-14 01:26:05.324] [I] [1-1] Build: 105b9f97
[2025-06-14 01:26:05.324] [I] [1-1] Level Zero: 1.11.0
[2025-06-14 01:26:05.324] [I] [1-1] xpumd core starts to initialize
[2025-06-14 01:26:05.324] [I] [1-1] initialize configuration
[2025-06-14 01:26:05.324] [I] [1-1] xpum mode: xpum
[2025-06-14 01:26:05.324] [I] [1-1] initialize datalogic
[2025-06-14 01:26:05.324] [I] [1-1] initialize device manager
[2025-06-14 01:26:05.351] [E] [1-1] GPUDeviceStub::init zeInit error: 78000001
[2025-06-14 01:26:05.351] [I] [1-1] GPUDeviceStub::checkInitDependency start
[2025-06-14 01:26:05.351] [I] [1-1] Environment variables check pass
[2025-06-14 01:26:05.357] [I] [1-1] Libraries check pass.
[2025-06-14 01:26:05.357] [I] [1-1] Permission check pass.
[2025-06-14 01:26:05.357] [I] [1-1] GPUDeviceStub::checkInitDependency done
[2025-06-14 01:26:05.357] [E] [1-1] xpumInit LevelZeroInitializationException
[2025-06-14 01:26:05.357] [E] [1-1] Failed to init xpum core: zeInit error
[2025-06-14 01:26:05.357] [E] [1-1] XPUM: Load xpum library failed! 40
[2025-06-14 01:26:05.357] [I] [1-1] XPUM: start XPUM RPC Server.
[2025-06-14 01:26:05.357] [I] [1-1] XPUM: start RPC server ...
[2025-06-14 01:26:05.359] [I] [1-1] XPUM: RPC server is listening at /tmp/xpum_p.sock
```

However, when I upgrade the dependencies from Intel's APT repo (changing the component to `jammy/unified`), everything works properly.

Reading through some of the issues ([1](https://github.com/intel/xpumanager/issues/17), [2](https://github.com/intel/xpumanager/issues/67)) on this repo, it seems like there are problems arising every few months from the container image being way out of date. **Please** start pushing every release to dockerhub (or another container registry, like github/ghcr.io) so that these issues can be avoided.