ARM Cortex processors are classified depending on the application for which it is intended as follows::
- Cortex-A (Application): For applications such as phones, tablets, smart TVs or mini PCs. They are processors capable of running an operating system, such as Linux. They can be found on boards such as the Raspberry Pi, Orange Pi and BeagleBone.
- Cortex-R (Real Time): For real-time applications such as data network equipment or automobiles.
- Cortex-M (Microcontroller): Simple, low-cost, low-power applications, such as IoT and embedded devices. Given the great variety of processors in this category, there is greater detail or subclassification as detailed below:
- M0: It is the simplest and smallest model, intended for low cost and low energy consumption applications.
- M0+: Improved version of the M0, with more power and lower energy consumption.
- M1: Designed specifically to be used with FPGAS.
- M3: A high-performance processor, adds more instructions to perform complex tasks. It includes division and multiplication instructions and a debug module that facilitates software debugging.
- M4: It is equivalent to M3 with the addition of signal processing (DSP) instructions.
- M7: It is a high-performance core, equivalent to the M4 but with better energy management. The internal data and instruction buses are 64 bits instead of 32 as in previous models.
- M23: It is equivalent to an M0+ with the addition of mathematical instructions and greater security.
- M33: Combines the best of the M4 and M23 models.
- M35P: It is an improvement of the M33 including more safety functions.
- M55: Includes functions that make it particularly useful in Artificial Intelligence (Machine Learning) applications.
- Cortex-X (Custom): More demanding applications such as desktop computers.
Special consideration for ARM Cortex processors are the NVIDIA GPU Computing processors intended for Artificial Intelligence.
GPU Computing is the way forward for high-performance computing (HPC) and data centers
The world's fastest supercomputer, as well as the most advanced systems in Europe and Japan, have been created with NVIDIA technology.
Summit, presented by the USA, is the smartest and most powerful supercomputer in the world, with more than 200 petaflops of high computing (HPC) and 3 exaops of AI. Summit merges HPC computing and AI with more than 27,000 GPUs with NVIDIA Volta Tensor Core to accelerate scientific discovery.
At Electrónica Embajadores you can find the modules that enable it:
NVIDIA Jetson Xavier NX is perfect for high-performance AI systems such as wearable medical devices, commercial robots, smart cameras, high-resolution sensors, automated optical inspection, smart factories and other integrated IoT systems. It brings supercomputer performance to the edge in a small form factor system-on-module (SOM). With up to 21 TOPS of accelerated computing it offers the power to run modern neural networks in parallel and process data from multiple high-resolution sensors, a requirement for fully AI systems.
NVIDIA Jetson AGX Xavier marks a new milestone in compute density, energy efficiency, and AI inference capabilities in edge devices. The next step in the evolution of next-generation intelligent machines with comprehensive autonomous capabilities.
Jetson AGX Xavier delivers up to 32 TOPS of AI performance, and the extensive set of AI tools and workflows from NVIDIA allows developers to quickly train and deploy neural networks.
As the world's first computer designed specifically for autonomous machines, the high-performance Jetson AGX Xavier can handle visual odometry algorithms, sensor fusion, localization and digital mapping, obstacle detection and path planning critical for next-generation robots. Get GPU workstation performance of 32 TeraOPS (TOPS) peak compute and unprecedented 750 Gbps high-speed I/O in a compact 100 x 87 mm form factor.
