Software and hardware requirements

The requirements for the base computers and operating system for the modules included in the Intellect Detector pack correspond to the same requirements for the Intellect (see Intellect Administrator's Guide). 

Each software module automatically determines which computing resources it can use — a central processing unit (CPU) or a graphics processing unit (GPU).

For the modules that use neuroanalysis (see General information on Neural analytics), the operation device can be set manually.

Requirements for the hardware platform when the neuroanalysis is used:

1. The neural network analytics supports the following devices: CPU, GPU (NVidia discrete GPUs, integrated Intel GPUs), VPU (Intel NCS, Intel HDDL).
   

   Note.

   In order to connect the Intel NCS, plug in the device to the USB port and make sure that Windows shows it with one of the following names: Movidius, Myriad X, VSC Loopback Device. 

   Intel NCS can be utilized with any computer compliant to the Intellect hardware requirements (see Requirements for base PCs).

   Attention!

   It is not recommended to use more than one Intel NCS device on the same Server.

   Attention!

   The neural network analytics functions are not available for AArch64 (ARM) architecture processors.

   
   

2. If CPU or Intel GPU is used for the neural network analytics operation, keep in mind that the following processors are supported: Intel Core generation 6 and later, Intel Xeon and Intel Pentium® processors N4200/5, N3350/5 or N3450/5 with Intel® HD Graphics https://software.intel.com/en-us/openvino-toolkit/hardware.
3. NVIDIA GeForce 1050 Ti GPU or newer. GPU requirements:
   
   1. at least 2 GB of memory;

   2. Compute Capability 3.0 or higher.

      Note.

      Please check the GPU's Compute Capability version on the manufacturer's web site.

      Attention!

      When using NVidia graphics cards, it is recommended to install the latest driver from the NVidia official web site.

A single neural network consumes 500 MB of video memory. Each neural network is distributed between several object trackers within the GPU. For example: to apply the neural fire detector and the neural smoke detector to any number of channels, you should use a 1 GB graphics card or higher. You can use multiple video cards in your system.