Axelera AI
Axelera AI is a Dutch fabless semiconductor company that designs D-IMC AI inference accelerators (Metis, Europa, Titania) and the Voyager SDK for edge-to-data-center deployment, serving enterprise customers across defense, industrial, robotics, retail, HPC, and space.
- Company typePrivate
- Founded2021
- HeadquartersEindhoven, Netherlands
- Headcount101–250
- GTM typeB2B
- OfferingHardware or Manufacturing
What Axelera AI does
Axelera AI is a Dutch fabless AI semiconductor company founded in July 2021 and headquartered at the High Tech Campus in Eindhoven, Netherlands. The company designs and sells purpose-built AI inference accelerators based on a proprietary Digital In-Memory Computing (D-IMC) architecture with RISC-V vector extensions, positioning itself as a European sovereign alternative to GPU-based AI inference. Its product portfolio spans three generations: the Metis AIPU (214 TOPS at approximately 10W) shipping today in M.2 and PCIe form factors, the Europa processor (629 TOPS, 3x–5x performance over Metis, samples expected Q2 2026) targeting automotive, robotics, and edge servers, and the Titania chiplet (expected 2027–2028) for HPC and data center deployments under the EuroHPC JU DARE Project. The Voyager SDK provides end-to-end model deployment built on Apache TVM, with a Model Zoo of 100+ pre-optimized models covering computer vision, NLP, and generative AI workloads.
The company monetizes through hardware sales (accelerator cards, compute boards, and evaluation systems) distributed via its own webstore, a global Partner Accelerator Network (Dell, Lenovo, Advantech, SECO, Aetina), and European distribution through Macnica. Supplementary revenue comes from Voyager SDK licensing and non-dilutive EU funding, including a €61.6M EuroHPC JU grant and participation in the CONVOLVE, CHASSIS, and RIGOLETTO EU projects. Axelera serves enterprise customers across defense and security, industrial manufacturing, robotics, retail, agritech, physical security, HPC, space and satellite systems, and smart cities, with named deployments at the European Space Agency, Bull, Telefonica, SECO, 2CRSI, and other partners. As of early 2026 the company reported over $450 million in cumulative funding and 500+ global customers, with an estimated ~$71M in annual revenue and a pipeline cited above $100M.
Axelera AI firmographics
Firmographics- Name
- Axelera AI
- Legal name
- Axelera AI B.V.
- Website
- https://axelera.ai
- Company type
- Private
- Founded year
- 2021
- Operating status
- Operating
- Headcount range
- 101–250 employees
- Short description
- Axelera AI is a Dutch fabless semiconductor company that designs D-IMC AI inference accelerators (Metis, Europa, Titania) and the Voyager SDK for edge-to-data-center deployment, serving enterprise customers across defense, industrial, robotics, retail, HPC, and space.
- Ownership category
- akta.pro rank
Axelera AI industry classification
Industry- Product category
- Edge AI Accelerators
- NAICS
- Computer and Peripheral Equipment Manufacturing (33411), Computer Terminal and Other Computer Peripheral Equipment Manufacturing (334118), Custom Computer Programming Services (541511), Search, Detection, Navigation, Guidance, Aeronautical, and Nautical System and Instrument Manufacturing (334511)
- SIC
- Electronic Computers (3571), Services-Prepackaged Software (7372), Services-Computer Integrated Systems Design (7373)
- akta.pro primary industry
- Edge AI Hardware & NPUs (SoCs, modules, accelerators) (HDAAAJAC)
- akta.pro secondary industries
- AI Accelerators (GPUs/TPUs/NPUs/ASICs) (HDAAAAAA), AI/Neural Processing Units (NPUs) & ML Accelerators (HDAHAJAF), AI Server Systems & HGX/Accelerator Platforms (HDAAAAAB), AI Compiler, Runtime & Kernel Optimization Software (CUDA/ROCm/XLA, graph compilers) (HDAAAAAI)
Keywords
Where Axelera AI is headquartered
LocationHeadquarters
- HQ city
- Eindhoven
- HQ country
- Netherlands
- HQ region
- Europe
Offices2 records
Markets served
Axelera AI business model
Business model- GTM type
- B2B
- Offering type
- Hardware or Manufacturing
- Cost components
- Technology or R&D, Personnel, Supply Chain, Marketing or Sales, Operations
Revenue model
- AI Accelerator Hardware Sales: Sale of Metis and Europa AI Processing Units (AIPUs) in M.2 and PCIe form factors, along with complete evaluation systems and development platforms. Manufacturing scale-up supported through partnerships with TSMC and Samsung.
- Software Licenses (Voyager SDK): Voyager SDK software stack licensed for deployment and optimization of AI models on Axelera hardware. Available via GitHub with open-source and commercial licensing terms. Software updates and technical support provided as additional services.
- EU Research Grants: Co-funded by European Union Horizon research programs and EuroHPC Joint Undertaking initiatives including CONVOLVE, DARE, RIGOLETTO, and CHASSIS projects, contributing to R&D funding.
- EU Grant (EuroHPC): Up to €61.6 million grant from EuroHPC JU and member states under the DARE Project for Titania chiplet development, expected to be deployed in HPC and data center solutions by 2027-2028.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Unit Pricing | Pay-as-you-go | Metis M.2 AI Accelerator Card — quad-core AIPU in compact M.2 form factor for embedded edge AI applications |
| Unit Pricing | Pay-as-you-go | Metis M.2 Max AI Accelerator Card — higher-performance quad-core AIPU in M.2 form factor for compute-intensive edge AI |
| Unit Pricing | Pay-as-you-go | Metis PCIe AI Accelerator Card — quad-core AIPU in PCIe form factor for desktop and edge server deployment |
| Unit Pricing | Pay-as-you-go | Metis PCIe AI Accelerator Card (4x quad-core) — high-density PCIe card for demanding edge AI workloads |
| Unit Pricing | Pay-as-you-go | Metis Compute Board — ARM-based RK3588 SOM with embedded Metis AIPU for complete edge AI development |
| Unit Pricing | Pay-as-you-go | Partner Evaluation Systems — pre-configured systems with Dell, Lenovo, Advantech, and SECO |
Go-to-market motion5 records
Distribution channels4 records
Marketing channels12 records
Axelera AI product offering
Product offeringCore offering
Axelera AI designs and sells purpose-built AI accelerator hardware (AI Processing Units, or AIPUs) based on its proprietary Digital In-Memory Computing (D-IMC) architecture and RISC-V technology for edge AI inference workloads. Its commercial lineup includes the Metis AIPU family (214 TOPS at ~10W) delivered as M.2 and PCIe accelerator cards, the Europa second-generation processor (629 TOPS) targeting automotive, robotics, and edge servers, and the Titania chiplet for HPC/data center deployments expected in 2027. The Voyager SDK, Model Zoo with 100+ pre-optimized AI models, and Metis Compute Board evaluation system round out the end-to-end offering for developers and system integrators.
Product overview
Axelera AI offers a three-generation AI inference hardware and software portfolio: Metis (first-gen, 214 TOPS) and Europa (second-gen, 629 TOPS) are edge AI processors in the AIPU family, with Titania (third-gen chiplet, expected 2027) targeting HPC and data centers. Hardware is available in M.2, PCIe, and compute board form factors. The Voyager SDK provides end-to-end software deployment with 100+ Model Zoo models. The platform uses proprietary Digital In-Memory Computing (D-IMC) architecture with RISC-V technology for power-efficient inference at the edge.
Differentiator
Problem solved
Functional benefit
Products and services
- Metis AI Processing Unit (AIPU) First-generation AI inference accelerator chip delivering 214 TOPS at approximately 10W using Digital In-Memory Computing (D-IMC) architecture, available in quad-core M.2 and PCIe configurations for embedded and edge server deployments.
- Metis M.2 AI Accelerator Card M.2 form factor edge AI accelerator card powered by a quad-core Metis AIPU, designed for embedded systems and space-constrained deployments.
- Metis M.2 Max AI Accelerator Card High-performance M.2 AI inference accelerator card that delivers PCIe-card-level performance in an M.2 form factor, designed for the most compute-intensive edge AI inference applications including LLMs.
- Metis PCIe AI Accelerator Card PCIe AI accelerator card powered by a quad-core Metis AIPU, available in single and 4x quad-core configurations for higher compute demands and capable of running 16 simultaneous YOLO V5 video streams per chip.
- Metis Compute Board Evaluation and development board with an ARM-based RK3588 processor integrated with the Metis AIPU, used for rapid prototyping and testing of edge AI inference applications.
- Voyager SDK End-to-end integrated software stack for deploying AI applications to Axelera Metis edge devices, featuring declarative YAML pipeline configuration, a Model Zoo of 100+ pre-optimized AI models, Apache TVM-based ML compiler with proprietary quantization, GStreamer-based preprocessing, and PyTorch compatibility. Available via GitHub with open-source and commercial licensing.
Quantifiable outcome
- Metis AIPU performs 214 TOPS at approximately 10W power consumption, representing industry-leading power efficiency for edge AI inference
- +4 more outcomes
Companies that use Axelera AI
Customer profileNamed customers6 records
Segments10 records
Ideal customer profiles4 records
Axelera AI technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
Integration7 records
AI capability8 records
Feature6 records
Axelera AI partnerships and signals
Strategic signalPartnerships
21 partnerships are on record, tiered minor, core and flagship.
- NEMA-MedBot-Miky (Community Project)minorCommunity assistive robotics platform integrating Axelera AI's Metis chip for neural inference, using OrangePi 5 Plus running ROS2 for orchestration with clinical validation planned via a medical partner in Brescia. Project code and documentation publicly available on GitHub.
- BullcoreBull (formerly Atos subsidiary, now independent following French State's acquisition) integrating Axelera AI's edge AI hardware accelerators into the BullSequana XH3500 AI-optimized supercomputer for Sweden's Mimer AI factory at Linköping University. Marked Bull's first major strategic announcement as an independent entity.
- Kudelski LabscoreIntegration of Kudelski Labs' Kudelski Secure Enclave (KSE3) into Axelera AI's Europa Edge AI platform during embedded world Germany, enhancing security features for edge AI applications in industrial automation and autonomous robotics. Demonstrated KLARQ autonomous robotic platform showcasing EU Cyber Resilience Act compliance.
- Multiverse ComputingcoreStrategic technology collaboration to deploy compressed AI models directly on Axelera's Metis and Europa edge computing platforms using Multiverse's CompactifAI technology that can shrink LLMs by up to 95% with only 2-3% precision loss. Joint commercialization program targets industrial, retail, mobility, defense, and smart city deployments. Supported by the European Innovation Council and framed around European technological sovereignty goals.
- DeGirumcoreTechnology partnership integrating DeGirum's computer vision, generative AI, and inferencing applications with Axelera AI's Metis Edge AI Processors. Provides Axelera customers one year of premium technical support from DeGirum, scalable application deployment, and remote hardware performance testing capabilities.
- Neuromorphic Computing NL (NC-NL)minorNew national coalition launched in the Netherlands advancing energy-efficient neuromorphic computing technology, bringing together universities, research organizations, and companies including Axelera AI, Innatera, Hoursec, and IMChip to address AI energy consumption challenges.
- European Space Agency (ESA)flagshipStrategic partnership to embed Metis AI inference accelerator into ESA's future European space infrastructure for real-time on-board satellite processing. Collaboration addresses radiation hardening challenges for decade-long satellite missions, with ESA's Φ-sat-2 nanosatellite serving as operational validation.
- European Space Agency (ESA) Φ-lab IrelandminorESA's national hub for space-enabled AI launched February 13, 2026, working alongside IBM Research Europe's TerraMind foundation model to provide software infrastructure for Axelera's space-rated hardware accelerators.
- ArteriscoreArteris licensed its FlexNoC 5 interconnect IP to Axelera AI for development of the Europa platform, enhancing high-performance AI inference at the edge through physical awareness technology that supports various AI frameworks with power and performance efficiency.
- SECOcoreSECO integrates Axelera Metis AIPU into its SOM-COMe-BT6-RK3588 module for automotive inspection applications, demonstrating real-time multi-stream video processing with advanced AI acceleration for industrial quality control.
- UltralyticscoreJoint demonstration partnership at CES 2026 showcasing YOLO model performance on Axelera's Metis platform. Ultralytics YOLO models are featured in Axelera's Voyager SDK Model Zoo for computer vision applications.
- AdvantechcoreAdvantech offers pre-configured evaluation systems combining Metis PCIe with MIC-770v3 and ARK-3534 embedded computers. Advantech CEO Joey Hsu cited Axelera's sovereign technology and long-term availability as key selection criteria for ESA missions.
- Dell TechnologiescoreDell offers Metis PCIe integrated with Dell Pro Slim Plus XE5 as a validated edge AI system, available through Dell's commercial sales channel and Axlvera's webstore.
- LenovocoreLenovo offers Metis PCIe with Lenovo ThinkStation P360 Ultra as a validated evaluation system for edge AI deployment.
- MacnicacoreMacnica as Axelera AI's first European distributor, bringing state-of-the-art AI acceleration technology to customers across Europe with industry-leading performance and efficiency for edge applications. Partnership reduces development complexity and time-to-market for European customers.
- CHASSIS (Chips JU Project)core€70.7M Chips Joint Undertaking project (November 2025 – October 2028) with 19 partners across 8 countries. Axelera contributes a high-performance AI accelerator chiplet built on D-IMC architecture for automotive applications (ADAS, autonomous driving), serving as consortium Risk Manager.
- RIGOLETTO (Chips JU Project)coreThree-year Chips JU project (2025–2028) building next-gen automotive computing platform on RISC-V. Consortium led by Infineon with Axelera, NXP, STMicroelectronics, Bosch. Axelera co-designing on-chip math features for efficient neural network processing and validating ultra-compact number formats.
- TelefonicacoreTelefonica enabling best-in-class AI solutions integrating seamlessly with global partners' platforms, advancing energy-efficient inference-focused AI technologies developed in Europe for sustainable and resilient AI deployments.
- 2CRSIcore2CRSI ensuring sovereign, predictable, Europe-based AI supply chain for resilient edge to HPC systems. Collaboration strengthens long-term availability across 2CRSI's portfolio from embedded platforms to large-scale datacenter deployments.
- DARE Consortium (EuroHPC JU)flagship€240 million EuroHPC Joint Undertaking project coordinated by Barcelona Supercomputing Center with 38 organizations. Axelera developing Titania AI inference chiplet as part of this initiative, awarded up to €61.6M funding for scalable HPC and AI chiplet development targeting 2027-2028 deployment.
- CONVOLVE (EU Horizon Project)coreEU Horizon Europe project with 18 academic and industry partners enforcing Europe's position in smart edge processor design and development, taking a holistic approach from application through circuit and device level. Axelera contributes to the full design stack.
Scale indicators12 records
Recent moves6 records
Expansion highlights6 records
Axelera AI competitors and assessment
Company assessmentDirect peers
- Hailo: Israel-based edge AI accelerator company shipping M.2/PCIe inference modules with a competing TOPS/W proposition for computer vision at the edge. Most direct comparable to Axelera's Metis line in target use cases, form factors, and customer segments.
- SiMa.ai: Edge AI inference chip startup focused on multimodal workloads at low power, also co-listed with Axelera in CRN's 2026 IoT 50. Competes head-to-head in industrial, automotive, and smart vision edge deployments.
- Mythic: Analog in-memory computing edge AI chip company targeting power-constrained inference. Mentioned alongside Axelera in CRN's 2026 IoT 50; competes in the same analog/mixed-signal alternative-architecture narrative for edge AI.
Emerging players
- Innatera: Dutch neuromorphic spiking-neural-network chip company; co-member with Axelera in the NC-NL neuromorphic computing coalition. Emerging player in ultra-low-power edge AI with overlapping Dutch deep-tech ecosystem and talent pool.
- Tenstorrent: RISC-V-based AI accelerator company building chiplet-based inference and training hardware. Directly comparable on RISC-V/open-architecture positioning and on the chiplet roadmap that Axelera is pursuing with Titania.
- Groq: AI inference silicon company focused on low-latency, high-throughput language model serving. Competes for inference workloads at scale, intersecting with Axelera's Europa/Titania ambitions in the data center tier.
Broad incumbents
- Qualcomm: Global mobile and edge AI platform incumbent with AI Engine on Snapdragon and IoT chipsets. Co-listed with Axelera in CRN's IoT 50; represents the largest incumbent threat in edge inference across smartphones, IoT, and automotive.
- Intel: Incumbent CPU/accelerator vendor with Movidius, Habana Labs, and OpenVINO stack serving edge and data center AI. Competes on full-stack software breadth and supply-chain scale; co-listed in CRN's 2026 IoT 50.
- NVIDIA: Dominant AI accelerator incumbent (Jetson for edge, Tensor Core GPUs for data center) with the deepest AI software stack (CUDA, TensorRT). The primary competitive benchmark on TOPS/W and developer mindshare for any AI chip company.
- AMD (Xilinx): Broad semiconductor incumbent with Versal adaptive SoCs, Ryzen AI, and Alveo accelerators covering edge to data center. Competes with Axelera in industrial, automotive (ADAS), and embedded vision segments.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Key risks6 records
Key highlights7 records
Customer concentration
Axelera AI social profiles
Digital presenceAxelera AI financial estimates
Financial estimateRevenue estimate
Valuation estimate
Axelera AI leadership team
Management profileNumber of profiles
Profiles10 records
Axelera AI funding detail
Funding detailFunding overview
Funding rounds10 records
Investors17 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Axelera AI M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Axelera AI
What does Axelera AI do?
Axelera AI designs and sells purpose-built AI accelerator hardware (AI Processing Units, or AIPUs) based on its proprietary Digital In-Memory Computing (D-IMC) architecture and RISC-V technology for edge AI inference workloads. Its commercial lineup includes the Metis AIPU family (214 TOPS at ~10W) delivered as M.2 and PCIe accelerator cards, the Europa second-generation processor (629 TOPS) targeting automotive, robotics, and edge servers, and the Titania chiplet for HPC/data center deployments expected in 2027. The Voyager SDK, Model Zoo with 100+ pre-optimized AI models, and Metis Compute Board evaluation system round out the end-to-end offering for developers and system integrators.
Is Axelera AI a public or private company?
Axelera AI is a private company. It is classified as venture growth investor backed and is currently operating.
When was Axelera AI founded?
Axelera AI was founded in 2021. It employs 101 to 250 people.
Where is Axelera AI based?
Axelera AI is headquartered in Eindhoven, Netherlands, in the Europe region.
How does Axelera AI make money?
Four revenue lines are on record. AI Accelerator Hardware Sales are the primary driver. The others are software Licenses (Voyager SDK), EU Research Grants and EU Grant (EuroHPC).
Who are Axelera AI's main competitors?
Direct peers on record are Hailo, SiMa.ai and Mythic. Emerging players are Innatera, Tenstorrent and Groq. Broad incumbents are Qualcomm, Intel, NVIDIA and AMD (Xilinx).
Does Axelera AI have an API?
No public API is recorded for Axelera AI.
What industry is Axelera AI in?
Axelera AI's product category is Edge AI Accelerators. Its primary akta.pro industry code is HDAAAJAC, Edge AI Hardware & NPUs (SoCs, modules, accelerators), with a secondary code of HDAAAAAA, AI Accelerators (GPUs/TPUs/NPUs/ASICs). Its NAICS code is 33411 and its SIC code is 3571.