Skip to content
Shop
Search icon
Shop
AI ACCELERATORS

Metis M.2 Inference Acceleration Cards

Axelera® AI’s M.2 cards, powered by Metis® AIPU, is perfect for constrained devices that require high-performance AI inference.

The M.2 and M.2 Max configurations allow customers to right-size their systems to best fit their business needs.

Metis M.2 Max

Metis M.2 Max brings the performance of a PCIe card to the M.2 form factor. Double the memory bandwidth, dramatically increased capacity up to 8 GB, a slimmer profile, advanced thermal management features, and enhanced security capabilities give you even more flexibility for demanding workloads at the edge.

 

M.2 AI Edge accelerator card. Powered by a quad-core Metis AIPU

Providing best-in-class performance, the M.2 AI accelerator of Axelera AI is the best solution for AI acceleration. Powered by a single Metis AIPU and containing 1 GB DRAM dedicated memory, it minimizes power consumption and simplifies integration. It opens the door for every application to benefit from AI processing. Providing a significant boost to any vision processing system with an NGFF (Next Generation Form Factor) M.2 socket.

Order
Download Product Brief
Benchmarks
M.2 AI Edge accelerator card.
M.2 AI Edge accelerator card.

Metis M2. Card Performance Benchmark

 FPS Measurements Across Various Networks(higher is better)Metis M.2 CardCompetitor 1Competitor 2YOLOv5s(640x640)YOLOv5m(640x640)YOLOv7(640x640)YOLOv8l(640x640)01002003004005006007008009001,00083232217614254315610064334241128104NetworkPerformance in Frames Per Second (FPS)
 

*Metis performance will keep improving over time.

*Competitor 1,2 are other AI accelerators. Data obtained from public sources as of April 2026.

 

M.2 Max AI accelerator card. Powered by a quad-core Metis AIPU

Providing breakthrough AI performance, Axelera AI’s M.2 Max card is the ideal solution to unlock LLM and VLM applications on edge and embedded devices. Powered by the Metis AIPU and available with up to 8 GB of dedicated memory, it maximizes efficiency while minimizing power consumption. It enables multi-camera vision inference with cascaded or parallel models, opening new possibilities for intelligent applications. With secure boot, extended temperature options, and effortless software integration through the Voyager® SDK, the M.2 Max delivers uncompromised performance and flexibility in a compact form factor.

Coming soon
Download Product Brief
Benchmarks
M.2 Max AI accelerator card.
M.2 Max AI accelerator card.

Metis M.2 Max Card Performance Benchmark

 

*M.2 Max performance data is PRELIMINARY.

*Competitor 1,2 are other AI accelerators. Data obtained from public sources as of April 2026.

Key Features

Rocket
High-performance M.2 card features Metis AIPU to enable state-of-the-art AI inference in small-footprint devices.
Network
A single board can run inference on multiple cameras as well as support many independent parallel neural networks.
Data
A wide range of end-to-end AI pipelines and models are available out of the box.
Integration
Hassle free evaluation and software integration thanks to Voyager® SDK.

 

Electronic
Uncompromised prediction accuracy thanks to advanced quantization tools.

 

Fast and Efficient AI inference hardware that does not break budgets.

Whether you are a computer vision system developer or integrator, software vendor or OEM, our AI acceleration hardware has been built to meet your needs. Delivering leading AI acceleration hardware in a range of industry accepted form factors supported by our easy-to-use software stack, Metis simplifies development, integration and deployment of AI inference acceleration.

Try Metis Today

Bram Verhoef, Vice President of Customer Engineering & Success at Axelera AI, demonstrates our latest edge AI and vision technologies and products at the 2024 Embedded Vision Summit. Specifically, Bram gives a product line overview and demonstrates the types of computer vision applications that Axelera AI’s Metis AIPU accelerates

See Metis in Action

24-live stream cameras

We showcase 24 concurrent live-streams, showing industry-leading multi-stream inference for video analytics, on a single Metis AI Processing Unit. Experience 24 live camera streams being processed using YoloV5 object detection based on a single Metis PCIe card.

Enabling a safer workplace

Together with our Partner Fogsphere we will highlight the power of AI will make the workplace safer. We will live demonstrate a Personal Protective Equipment (PPE) use-case, showing the capability for real-time monitoring of safety gear usage, such as helmets and vests. .