Sky-MXM-A2000 GPU-accelerated solution

Share this on social media:

Advantech, a leading provider of industrial edge computing platforms, provides a new series of solutions for embedded computing powered by the latest NVIDIA GPUs. The newest SKY-MXM-A2000 industrial solution will be released shortly.

Built on the NVIDIA® Ampere architecture, providing the highest performance

This industrial-grade GPU-accelerated solution is the first embedded MXM-based module built on the NVIDIA Ampere architecture, which delivers the latest RTX technology. It features real-time ray tracing, accelerated AI, advanced graphics, and high-performance computing capabilities, which is important for embedded computing solutions in enterprise deployments.

Supporting both RT and Tensor Cores for graphics and intensive-compute tasks

SKY-MXM-A2000 is equipped with GDDR6 graphics memory with error correction code (ECC) support for error-free computing. Most importantly, it uses an NVIDIA RTX™ A2000 embedded GPU, which combines 20 second-generation RT Cores, 80 third-generation Tensor Cores, and up to 2,560 next-generation CUDA® cores. SKY-MXM-A2000 fulfills increasing demands for processing power, graphics, and intensive-compute tasks for designers, engineers, and scientists, to support their application development in industries including auto-optical inspection (AOI), medical care, transportation, and gaming.

Compact design with high-end clock rates and low power consumption

The clock rates of SKY-MXM-A2000 range from 1,215 MHz (base) to 1,687 MHz (boost), which can power high performing workstations. Moreover, its compact size and ruggedized design with a small form factor of only 82 x 70 mm allows it to fit into limited spaces — with low power consumption (up to 60W). Its low thermal emissions make it suitable for intelligent systems like those in surgery, autonomous driving, and driver-assisted applications. With 0~55 °C standard operating temperatures, SKY-MXM-A2000 can withstand the demands of harsh industrial environments and ensure workstation stability and durability.

Industrial GPU Solutions with MS-Hybrid model support

SKY-MXM-A2000 supports two optional discrete and MS-Hybrid modes. As a pure accelerator card, it can focus on big data processing in MS-Hybrid mode, which is mainly configured for notebooks and high-end GPUs. In discrete mode, it supports a maximum of four DisplayPort signal outputs. In addition, Advantech's strict revision and engineering change control, as well as its 5-year product support, will reassure end customers.

Key Features

• NVIDIA RTX A2000 with MXM 3.1 TYPE A form factor (82 x 70 mm)

• Up to 2,560 CUDA cores, 20 RT Cores, 80 Tensor Cores (NVIDIA Ampere GPU architecture cores)

• GDDR6 4GB memory, 128-bit, bandwidth 192 GB/s

• Up to 4x DisplayPort 1.4a digital video outputs

• Long life cycle, supports 5 years' availability

Advantech, an Elite Partner in the NVIDIA Partner Network, provides edge systems and solutions applied in the fields of industrial, manufacturing, and robotics. Advantech is dedicated to developing cutting-edge GPUs like SKY-MXM-T1000 (built on the NVIDIA Turing™ GPU architecture with MXM 3.1 TYPE A, GDDR6 4GB and 896 CUDA cores) and up-comping SKY-MXM-RTX3000 (built on the NVIDIA Turing GPU architecture with MXM 3.1 TYPE B, GDDR6 6GB, 1920 CUDA cores, 30 RT Cores and 240 Tensor Cores).

Stephen Su shares how Arm’s Cortex-M85 processor and software ecosystem can be leveraged to overcome the constraints of microcontroller unit platforms

16 March 2023

Stephen Su shares how Arm’s Cortex-M85 processor and software ecosystem can be leveraged to overcome the constraints of microcontroller unit platforms

16 March 2023

The FSM-IMX570 Devkit provides a simple, coherent framework for quickly developing a working prototype of an indirect time-of-flight embedded vision system

15 March 2023

Maharajan Veerabahu, co-founder at E-Con Systems, shares the industries in which embedded vision and AI integration shows great promise

15 March 2023

Jan-Erik Schmitt, of Vision Components, charts the rise of embedded vision and the many advantages the technology has to offer

14 March 2023