Crafted in Europe, Trusted Worldwide.

Purpose-built AI accelerators that keep your data, your compute, and your analytics under your control. High performance. Low power. No compromise on sovereignty.

Our Technology

Digital in‑memory computing.

Compute inside SRAM. The data stops travelling, responses are faster & the watts stop burning.

  • UP TO

    5x

    performance per dollar

    Cost efficiency with no performance compromise

  • 15

    TOPS per watt

    Demanding workloads without the power, the cost or the heat.

  • 99.9%

    relative accuracy

    Post-training quantization preserves FP32-equivalent accuracy with no retraining

Metis AIPU opening into its layers
The chip multiplies into four
Four chips dock onto the PCIe card
Digital in-memory computing

The array becomes the chip

Inside our cores

Multi-core architecture, with all the compute you need

No compromise

99.9% accuracy on the most demanding AI workloads

Scaling

Chips that work together

A card for every use-case

Embedded to edge to server

Software Development Kit

One platform for any model

Scale

Deploy from robot to rack

From silicon to data center

One chip becomes the whole stack.

Chip, card, server, rack, data center: one architecture at every scale.

OUR PRODUCT Line

Three ways to deploy.

  • Embedded

    AI in the machine: M.2 modules and single-board computers

  • Edge

    AI next to your data: PCIe cards for edge servers and industrial PCs

  • Server

    AI at data center scale: multi-AIPU cards that bring inference density to racks and supercomputers.

Benchmarks

Ahead on every benchmark

Detection

YOLOv8l, 640 × 640, frames per second

Axelera Metis

0 FPS

Competitor

0 FPS
0x
Throughput

Ultralytics YOLOv7, 640 × 640, frames per second

Axelera Metis

0 FPS

Competitor

0 FPS
0x
Cost efficiency

YOLOv11l, frames per second per dollar

Axelera Metis

0 FPS per $

Competitor

0 FPS per $
0x
Energy efficiency

YOLOv11l, frames per second per watt

Axelera Metis

0 FPS/W

Competitor

0 FPS/W
0x
Video

Europa, YOLOv8l 640 × 640, frames per second

Axelera Europa

0 FPS

Competitor

0 FPS
0x
Reasoning

Qwen3-30B-A3B, tokens per second, batch 1

Axelera Europa

0 tokens/s

Competitor

0 tokens/s
0x
Examples

Axelera in action

Real-use cases running on Axelera silicon

  • TILED INFERENCE high-resolution scenes sliced into sections, so small objects stay detectable

  • SEGMENTATION AND DETECTION self-checkout: shopper, basket, bags and products masked and labelled, 18.6 to 22.6 fps end to end

  • PPE DETECTION hard-hat compliance across a wall of 24 live streams, with the host CPU at 4%

  • LLM, illustration an on-device language model answering across twelve camera streams

Embedded · Edge · Server

From Chip to Complete System

Choose how you deploy.

Solutions

One platform, multiple end-users

  • Industrial Manufacturing

    Edge AI for machine vision and Industry 4.0 equipment, raising productivity in production lines and warehouses.

  • Retail

    Real-time analytics across every customer touchpoint, so inventory control is easier and shrinkage reduces.

  • Security

    Video analytics that turns a camera from a video source into a smart sensor for investigation and risk prevention.

  • Healthcare

    High-performance AI at the edge keeps patient data safer while filling critical gaps in healthcare systems.

  • Smart Cities

    Technology and data-driven solutions that make cities more efficient, sustainable and responsive to residents.

  • Agriculture

    Better crop yields, optimized nutrients in the soil, and the right amount of water reaching the roots.

  • Space

    Autonomous rovers navigating distant planets, and AI-powered satellites monitoring Earth and deep space.

  • Computer Vision

    Vision moved to where the data is created: real-time processing, less data movement, scale into remote sites.

  • Robotics

    Robots that enter non-permissive environments and make real-time decisions with no internet connection.

  • High Performance Compute

    Digital In-Memory Compute scales up: the Titania chiplet serves the high performance compute market directly.

Silicon

Three generations, one architecture

One architecture that spans embedded to rack-scale deployments

Shipping
Metis Die 640

Metis®

Quad-core AIPU

0 TOPS per chip
  • Typical application power 3.5-9 W
  • Memory up to 16 GB per chip
  • LLM capacity up to 8B parameters
  • YOLOv8l 181 FPS
  • ResNet-50 up to 3,200 FPS
Next generation
Europa-AIPU-web

Europa™

Double the cores, sixteen vector units, pre/post-processing on die.

0 TOPS per chip
  • TDP 45 W per card
  • Typical power 30-40 W
  • Memory up to 64 GB per chip
  • LLM capacity up to 32B parameters
  • Qwen3-30B-A3B 53.5 tokens/s
Announced
Titania

Titania™

A chiplet that scales the architecture into HPC.

From model to production, powered by the Voyager® SDK

Everything you need to build, faster.

One toolchain, every deployment: describe your pipeline in YAML and the Voyager SDK compiles, quantizes and runs it on any Axelera product line, with no retraining required.

Voyager SDK

Voyager SDK: the end-to-end software stack

Purpose-built for AI inference: a declarative YAML pipeline language, 100+ supported models through one toolchain, and API compatibility with industry standards. From idea to deployment quickly, with deep customization when you need it.

Axelera Wingman

Axelera Wingman: the agentic AI platform

Purpose-built for Axelera hardware with deep, curated knowledge of the SDK and documentation: Axelera Wingman builds, debugs and optimizes inference pipelines from natural language, at 20 to 30% higher accuracy than general-purpose AI coding tools.

Describe the workflow in plain language; Voyager Wingman creates and deploys it on device.
The running pipeline, inspected on real Metis silicon without leaving the desk.
Partner Accelerator Network

Partner Accelerator Network

From hardware to software to distribution, Axelera partners with leading companies to deliver complete AI solutions.

Advantech
Aetina
AIKO
AI Tech
AizoOn
Almawave
Archetipi Digitali
Arduino
ASRock
ASTUTE
Axiomtek
BCD Intelligent Solutions
Cidev
CandT Solutions
Dell Technologies
Digica
Digeiz
DOD Solution
Euro Composant
Innowise
Kudelski Lab
Lenovo
Macnica
Malogica Group
Marubun
Melchioni
Meridian Data Labs
MicroIP
Micron
MCM Japan
Nisko Projects
OEC Okaya Electrnics
Open Nebula
ORA Computing
Prodrive technologies
RidgeRun AI
Rinf.Tech
Robovision
Rutronik
SECO
ServerDirect
SmartStation
Spanidea
Synthera
Ultralytics
WebOccult
Xebia

What do you need?

  • Not sure which card?

    Find the right card with Voyager Wingman.

    Describe the workload, the box and the power constraints. Voyager Wingman knows the line-up.

  • Ready to test

    Go straight to the shop.

    Embedded M.2 modules, compute boards, and Edge and Server PCIe cards, ready to order today.

  • Talk to us

    Rather talk to a human?

    Contact sales and a person comes back to you.