Disclaimer: The following recommendations do not prevent thorough testing before deployment of our solution, as any variables may impact VidiReports performances.



Minimum System Specs

 

  • Hard drive: 1GB of free space for the software and database

  • Memory: 4GB RAM

  • CPU: see information below 

  • GPU: Vidireports does not need a graphic card

  • Camera: any USB webcam supporting the UVC norm (i.e. most mid and high range ones), as well as most IP cameras (in MJPEG or H264 mode). More details here.

  • OS:

    • Windows® 7 / Windows® 8 / Windows® 10 64-bit

    • Linux Ubuntu 14 or later


CPU requirements by edition


For new deployments 


We recommend an Intel I3, I5 or I7 CPU from 8th generation onward (CPU model number should start by 8 or higher, e.g. i7-8xxx, i5-9xxx, i310xxx). You will be able to run all editions of VidiReports on a recent Intel-based machine with video high resolution. 


Quividi also proposes its Pro version on ARM dual and quad core, from the A7 architecture and above (and the Android 7 OS). Note however that VidiReports can only run ARM with CPUs that support the NEON instruction set.


At last, please note that Quividi runs on LG WebOS but not on Samsung SoC screens (Tizen) at this stage.


For existing deployments


The table below highlights a few CPU models and the editions they can run. An exhaustive table is also attached below with resolution they can achieve.



Indoor installations:

 

Outdoor installations:


For outdoor deployments with vehicles detection, we recommend an Intel I5 or I7 CPU from 8th generation onward with 8go of RAM. 


Note: 

- On the Atom family (including Celeron), 3 physical cores must be reserved for Vidireports for anything involving Footfall. The Atom family cannot run VidiReports Vehicles.

- The above benchmark reflects the specific CPU models. The Intel i3, i5 and i7 range each comes with many variations and the performances will vary on the specific generation, clock, nb of threads of the selected processor. The ones mentioned in that table are relatively old and entry level.



Maximum camera resolution depending on CPU


One key aspect is whether you'll be dedicating one or more cores to VidiReports on the processor. The more cores VidiReports can use, the larger the image it can process. Remember however, that some CPU must generally be left aside for other systems running on the machine, such as a CMS application, and that VidiReports does not make use of the GPU resources. 


(click on the image to enlarge)




Note that the larger the video resolution, the longer the face detection distance will be, but the higher the CPU requirement will be. Also take into consideration the presence of a CMS, which will require additional CPU.


Also, adding more instances of Vidireports (ie analyze images from 2 different cameras on the same PC), will require more CPU.



Customized benchmark 


You can find more instructions on how to qualify a specific machine with our benchmarking tool here


If you need to calculate the maximum detection distance, the next article will help you do it.