AI Inference Server GPU accelerated Hardware Requirements

AI Inference Server GPU accelerated application requires IPC BX-59A device (at least) with GPU.

Demands	Value
App Version (latest)	2.6.0
Hard disk memory	4.4 GB
RAM memory (max.)	25.2 GB
Reserved memory for Python runtime	6 GiB (max 12 GiB)
Reserved memory for External (GPU) runtime	6 GiB (max 16 GiB)
CPU-Core demand (min.)	1 Core
CPU-Core demand (max.)	Unlimited
Processor Architecture	X86-64
Industrial Edge Management Version (downward compatibility)	Edge Management version ≥ 1.15.9, IED-OS version: simatic-ipc-ied-os-3.1.0-19-x86-64, IE Runtime version: 1.24.1-3
Docker Compose	≥ latest
Docker Engine Version	≥ 20.10.24
Kernel Version	≥ 6.1
Software dependencies	Databus v3.2.3 , Optional: External Databus v2.1.0 Optional: SIMATIC S7 Connector v2.3.2-27, Optional: SIMATIC S7+ Connector v2.2.0, Optional: OPC-UA Connector v2.4.2-11, Optional: PROFINET IO Connector v2.1.2, Optional: Vision Connector version 1.2.2, Optional: Basler Vision Connector 1.0.0+20240126.8
Hardware prerequisite	1 network interface
Protocols	-
Required Domain Access	-
Ports	IPv4 & IPv6 / HTTPS / Port 443 / Egress/Ingress
Publicly accessible	No
Further Interfaces	HTTPS – Rest API
User Interface	Yes
UI is Mobile optimized	No
Browser (optimized for)	Preferred: Chrome
UI Display Target Size	24”
UI Language	Default: English, German, Chinese
User Documentation Language	Default: English, German, Chinese
Tested on Industrial Edge Device Type & Version	IPC BX-59A (minimum: 32 GB RAM) with GPU, Virtual IED (version: ievd-1.24.2-1-a)
Supported IPC HW	IPC BX-59A (minimum: 32 GB RAM) with GPU, Virtual IED (version: ievd-1.24.2-1-a)
Excluded platform	none
App Labels	none

Notice

Only one AI Inference Server application instance (from the above list) can be installed on an edge device. It means you can install MLFB 6AV2170-0LA10-0AA0 (AI Inference Server) or MLFB 6AV2170- 0LA10-1AA0 (AI Inference Server – 3 pipelines) or MLFB 6AV2170-0LA11-0AA0 (AI Inference Server GPU & Vision – 1 pipeline) on a single edge device. You cannot install e.g. MLFB 6AV2170-0LA11-0AA0 (AI Inference Server GPU & Vision – 1 pipeline) and MLFB 6AV2170-0LA10-1AA0 (AI Inference Server – 3 pipelines) applications on a single device.

CAUTION

AI Inference Server should not be used in mission-critical scenarios with high risks (i.e. development, construction, maintenance or operation of systems, the failure of which could lead to a life-threatening situation or to catastrophic damage ("critical application")).** Examples of critical applications: Use in avionics, navigation, autonomous vehicle applications, AI solutions for automotive products, the military, medicine, life support or other life- critical applications.