Page History

When you use neural analytics (see see General information on neural analytics), the hardware requirements are the following:

Some
1. Take into account the limitations of NVIDIA SDK
specific features allow
2. . The neural analytics
to work only on Windows Server 2019 and Windows 10. 

Neural analytics can work on CPU, Nvidia GPU, VPU (Intel NCS, Intel HDDL).

Note

title	Attention!

3. can operate on the GPU only in the operating systems specified in the Operating system requirements page.
To connect Intel NCS, insert the device into a USB port and make sure that it is recognized by Windows OS as a USB device with one of the following names:
4. VPU requirements (Intel NCS):
   1. Connect a device to the computer USB port.
   2. Make sure that the Windows system identifies a device as
5. 1. Movidius, Myriad X, or VSC Loopback Device.
You can use Intel NCS on any computer that meets Axxon One hardware requirements (see Hardware requirements). We do not recommend using
6. 2. Use it on the computers that meet the Hardware requirements.
   3. Apply no more than one Intel NCS device
per Server.You can use several Intel HDDL devices
7. 1. on
one Server if they have the same revisions.
8. For Intel HDDL to work correctly with AMD processors, pre-install the OpenVINO™ toolkit version 2019.3.379 (see docs.openvino.ai).

If CPU or Intel GPU is used to run analytics, the following requirements must be met:

9. 1. the server.
Supported CPUs:

6th Generation Intel® Core™ processors and higher.

Note
title Attention!
When using a CPU lower than 6th Generation Intel® Core™ processors, the operation of the detection tools isn't guaranteed.

10. Intel® Xeon® v5 family and Intel® Xeon® v6 family.
11. Intel® Movidius™ Neural Compute Stick.
12. Intel® Neural Compute Stick 2.
Intel®
13. Supported CPUs and requirements:
    

    CPU type	Supported generations/models	Key requirements and conditions
    Intel® Core™	6th generation (Skylake) and higher	Support for the AVX2 or AVX512 instruction set (see ark.intel.com) We don't guarantee the detector operation on CPUs lower than 6th generation
    Intel® Xeon®	Families v5 (Broadwell) and v6 (Skylake)	Support for the AVX2 or AVX512 instruction set (see ark.intel.com). CPU compatibility with the OpenVINO toolkit version that you use (see software.intel.com)
    Intel VPU devices	Movidius™ Neural Compute Stick, Neural Compute Stick 2,

14. Vision Accelerator Design with Intel® Movidius™ VPUs

.
OpenVino must support the Intel CPU

16. Support from the OpenVINO toolkit version that you use (see software.intel.com)

.
17. CPU must support the AVX2 or AVX512 instruction set (see ark.intel.com).
NVIDIA GeForce 1050 Ti graphics card or higher. Requirements:
18. At least 2 GB of memory;
19. CUDA 11.1–11.4;

Compute Capability 3.5–8.6.

Note

title	Attention!

21. Requirements for NVIDIA GPU: 
    

    Parameter	Minimum/recommended requirements
    GPU model	NVIDIA GeForce GTX 1050 and later (for Detector Pack 3.14–3.15.1) NVIDIA GeForce GTX 1650 and later (for Detector Pack 3.15.2 and later)
    Video memory space	At least 2 GB. For the accurate calculation, we recommend using AxxonSoft Platform Calculator (see AxxonSoft Platform Calculator documentation)
    CUDA version	11.1–11.8 (for Detector Pack 3.14–3.15.1) 12.8.1 (for Detector Pack 3.15.2 and later)
    Compute Capability	6.0–9.0 (for Detector Pack 3.14–3.15.1) 7.5–12.0 (for Detector Pack 3.15.2 and later) You can check the version here or on the manufacturer's website
    Supported architectures	Kepler (late

CUDA is compatible with generations of computers with the following architectures: Kepler (partially

22. ), Maxwell, Pascal, Volta, Turing, Ampere, Ada Lovelace (partially)

(see CUDA).
23. You can check Compute Capability GPU version on the manufacturer's website.
24. If you use NVIDIA graphics cards, make sure to download the latest drivers from the manufacturer's website.

When you use a graphics card, a single neural network requires 500 MB of video memory, except for the neural network for Face Detection (see Hardware requirements for the Face Detection TV and its sub-tools) and License Plate Recognition RR (see Hardware requirements for License Plate Recognition RR and Vehicle Recognition RR). For example: a neural fire detection tool and a neural smoke detection tool, both with unlimited number of channels, require a 1 GB graphics card or higher. You can use multiple graphics cards in your system.

...

...

1. (for Detector Pack 3.14–3.15.1) Turing, Ampere, Ada Lovelace (partially), Hopper, and Blackwell (for Detector Pack 3.15.2 and later) In Windows OS, starting with the Detector Pack 3.15.2 version, the support for Hopper and Blackwell is added for the following detectors: Neural tracker Neural counter Human pose detector Equipment detector Fire detector Smoke detector Neural classifier Barcode detector Water level detector  Stopped object detector Privacy masking Person-based privacy masking
   Drivers	We recommend installing the latest driver version from the official website
   Configuration	You can use up to four GPUs on one server

2. For the correct operation, each detector must meet individualspecificrequirements (see Detectors) to:
   • Hardware platform,
   • Video stream and scene,
   • Image characteristics.