Computers with discrete GPU and AI

Basically, three systems are used here:

The NVIDIA GeForce are classic graphics cards, which in addition to the GPU for processing and calculating image models, also offer connection options for monitors. There are models in different performance classes and dimensions, which are not only used in professional applications, but are also widely used in private applications. The computers listed here are optimized for use with certain NVIDIA GeForce cards and are delivered with them installed.
The NVIDIA Jetson family of AI platforms, on the other hand, are complete embedded systems for the realization of autonomous machines. Each Jetson module contains a complete System-on-Module (SoM) including CPU, GPU, PMIC, DRAM and Flash memory. This enables the development of compact, fanless AI systems with moderate performance.
With the Movidius™ Myriad™ X Vision Processing Units (VPU) Intel offers a basis for the self development of AI accelerator cards. Thanks to an architecture that minimizes data movements, Movidius VPUs achieve a balanced ratio of energy efficiency and computing power, which is often most relevant in industrial applications, where energy efficiency and low power dissipation are important factors. In this section you will find Embedded Box PCs equipped with accelerator cards based on Movidius Myriad X VPU.

This provides the user with various AI solution approaches to fulfill the tasks set according to requirements and the application environment, whether in transportation as part of driver assistance systems, vehicle interior monitoring or autonomous driving, or stationary in visual retail, autonomous control of machines and processes, traffic monitoring and much more.