Aetina CoreEdge – MXM AI Module
Aetina CoreEdge provides a variety of small form factor (SFF) modules with compact architecture, targeting applications that usually require high-performance computing in constrained spaces. These SFF modules include the Mobile PCIe Module (MXM) and M.2 Module, along with our proprietary evaluation Expansion Kit and one-stop thermal service. The Mobile PCI Express Module (MXM) features a compact commercial off-the-shelf (COTS) solution. Aetina leverages cutting-edge NVIDIA GPUs and AI ASIC processors, delivering unparalleled power efficiency and high-level computing capabilities. It is ideal for embedded systems that require high performance along with size, weight, and power (SWaP) optimization for edge AI applications.
NVIDIA RTX PRO Blackwell MXM – Coming Soon, Ready to Accelerate AI!
Coming in Q4 2025：Get ready for the latest MXM module powered by NVIDIA RTX Pro Blackwell! With the latest NVIDIA Blackwell architecture, it combines ultra-fast GDDR7 memory, 5th-gen Tensor Cores, and 4th-gen RT Cores to push AI performance to new heights. From entry-level to cutting-edge AI, the NVIDIA RTX Pro 500 to 5000 Blackwell models are built to accelerate your AI innovations.
Purpose-Built for Embedded Vision AI
With a seamless mix of services and advantages in MXM modules, Aetina equips edge AI to conquer demanding environments and unique needs.
Slim and Space-saving
- Small footprint and light weight
- Mounted flat to save mechanism space
Golden Finger 30µ"
- Extra protection from scratch and damage
- Ensure stable and quality signals
Extended Temperature
- Option temperature support : -20~70°C, -40~85°C
- Individual validation before shipping
Conformal Coating
- Protection against dust, moisture and corrosion
- Improve MTBF
CUDA Computing
- Up to thousands of CUDA cores
- Optimized parallel computing
Visual Computing
- Dedicated for AI acceleration
- Real-time image processing
Multi Displays
- 3840x2160 resolution
- DP++, HDMI outputs
Configurable TDP
- Power cap customization
- Meet specific usage scenario
As the dimensions of small form factor modules are inconsistent in the market, it’s inconvenient for users to customize heat dissipation design to avoid the occurrence of high temperature failure especially when developing high computing performance applications in smart medical, factory automation, and so on. To improve the situation, Aetina decided to roll out one-stop thermal service, including standard heat spreaders, optional semi sink, and customized cooler.
- Increase the area of dissipation for each IC component
- Handle temperature overheat issue not merely for GPU but memory as well
- High-watt heat spreader is made of copper, and low-watt one is aluminum
- Easy assembly
- Save extra jiq cost
Developer Evaluation Kits for MXM
In the early stages of AI deep learning project development, developers often spend significant cost and time building test systems to confirm performance, specifications and compatibility with related peripheral. To address the challenge, Aetina’s expansion kits provide onboard high-performance computing slot for preliminary evaluation across various applications, such as deep learning and computer vision.
- Support PCI Express Gen4 x16 slot to host
- Support 4x HDMI display output per board
- Compatible with Aetina's Embedded MXM modules, including Turing, Ampere, and Ada series
