# Europa AIPU by Axelera AI

Canonical page: https://technooptimist.io/computing/europa
Frontier: [Computing](https://technooptimist.io/computing.md) · Field: Inference chips · Company: [Axelera AI](https://technooptimist.io/companies/axelera-ai.md)
Stage: Deploy (4 of 5: Research → Proto → Pilot → Deploy → Scaling)
Updated: last update 2026-09-17, facts checked 2026-10-11
Goal: Run modern AI models, including language models, on devices and local servers at a fraction of a GPU's power and cost, so robots, cameras and factories can use AI without sending data to a cloud.
Status: Shipping as of Sep 2026, validated in Dell and Supermicro systems; Titania data-centre chiplet due 2027-28.
Next milestone: Titania data-centre chiplet, target 2028

## Milestones

| Status | Milestone | Date | Slip |
| --- | --- | --- | --- |
| Done | Metis announced: 214 TOPS at 15 TOPS/W | 15 Dec 2022 |   |
| Done | Up to €61.6M EuroHPC grant for Titania chiplet | 6 Mar 2025 |   |
| Done | Europa architecture unveiled (629 TOPS) | Oct 2025 |   |
| Done | $250M+ Series C | Feb 2026 |   |
| Done | Europa products shipping; Dell, Supermicro systems | Sep 2026 |   |
| Current | Titania data-centre chiplet | 2028 |   |

## Upcoming events

- **2028**: Titania data-centre chiplet

## Most important updates

- **17 Sep 2026**: Axelera's Europa chips ship in Dell and Supermicro systems as customers pass 600. Products based on Europa are now shipping and work with selected Dell and Supermicro servers, moving Axelera from edge devices into enterprise servers. ([our story](https://technooptimist.io/stories/axelera-s-europa-chips-ship-in-dell-and-supermicro-systems-as-customers-8e454ffe.md) · [original source](https://ioplus.nl/en/posts/axelera-ai-takes-on-nvidia-with-its-new-europa-chip))
- **Feb 2026**: Axelera AI raises over $250M Series C, the largest investment in an EU AI chip company. Axelera AI raised a $250M Series C led by Innovation Industries, with BlackRock and SiteGround Capital joining existing backers. ([our story](https://technooptimist.io/stories/axelera-ai-raises-over-250m-series-c-the-largest-investment-in-an-eu-ai-881d01f5.md) · [original source](https://www.osborneclarke.com/news/osborne-clarke-has-advised-bitfury-top-holdco-bvon-250-million-series-c-investment-axelera-ai))
- **24 Oct 2025**: Axelera unveils Europa: 629 TOPS, 128 MB on-chip memory and 16 RISC-V cores. Axelera announced its second-generation Europa chip with eight second-generation AI cores delivering up to 629 TOPS, plus 16 RISC-V vector processors, 128 MB of on-chip SRAM and 200 GB/s of LPDDR5 bandwidth. ([our story](https://technooptimist.io/stories/axelera-unveils-europa-629-tops-128-mb-on-chip-memory-and-16-risc-v-052fb0d6.md) · [original source](https://convergedigest.com/axelera-launches-europa-aipu-with-629-tops/))
- **6 Mar 2025**: Axelera AI gets up to €61.6M from EuroHPC to build Titania, a data-centre inference chiplet. Through the EU's DARE RISC-V project, Axelera received up to €61.6M to develop Titania, an inference chiplet combining its in-memory compute with RISC-V cores, with several chiplets packaged together. ([our story](https://technooptimist.io/stories/axelera-ai-gets-up-to-61-6m-from-eurohpc-to-build-titania-a-data-centre-c89ac1e6.md) · [original source](https://www.dutchnews.nl/businesswire/axelera-ai-secures-up-to-e61-6-million-grant-to-develop-scalable-ai-chiplet-for-high-performance-computing/))
- **15 Dec 2022**: Axelera AI announces Metis, a 214 TOPS edge AI chip built on digital in-memory computing. Axelera AI unveiled its first chip, Metis, a quad-core inference processor whose cores multiply inside SRAM memory. ([our story](https://technooptimist.io/stories/axelera-ai-announces-metis-a-214-tops-edge-ai-chip-built-on-digital-in-b0bdbf10.md) · [original source](https://axelera.ai/axelera-ai-announces-metis-ai-platform))

## Current obstacles

- **Model support**: Each new model family must be compiled and tuned for Axelera's chips before customers can switch from NVIDIA Jetson-class parts.
- **Customers train elsewhere**: Customers train on NVIDIA and run inference on Axelera, so each deployment means moving models between two toolchains.

## Physics limits

- **SRAM is small**: In-memory compute uses SRAM, which stores far fewer bits per area than DRAM. Large language models spill into slower external memory, where bandwidth sets the speed.
- **Low precision has a cost**: Peak figures assume 8-bit integers. Models need careful quantisation to keep accuracy, and some layers still need higher precision at lower speed.

## How it works

### Bottleneck: Moving weights costs the energy

Running a neural network is mostly multiply-and-add on stored weights; fetching those weights from memory burns more energy than the maths.

### In-memory: Compute where weights sit

Each AI core does the multiplications inside SRAM arrays holding the weights, digitally, so it keeps exact results without retraining.

### Helpers: RISC-V cores and video decode

Sixteen RISC-V vector cores handle the other steps, and a built-in video decoder takes several 4K camera streams straight in.

### Scale: One to four chips per card

PCIe cards carry up to four Europa chips and 256 GB of memory, enough for larger language models in a server.

## Spec sheet

| Spec | Europa AIPU | Metis AIPU | Provenance |
| --- | --- | --- | --- |
| Peak compute (INT8) | 629 TOPS | 214 TOPS (Metis) | reported |
| AI cores | 8 second-gen D-IMC cores | 4 (Metis) | reported |
| RISC-V vector cores | 16 (4,915 GOPS) |   | reported |
| On-chip L2 SRAM | 128 MB | 52 MiB (Metis) | reported |
| Memory bandwidth | 200 GB/s (256-bit LPDDR5) |   | reported |
| PCIe card memory | 16 GB (1 chip) to 256 GB (4 chips) |   | reported |
| Customers | More than 600 |   | reported |

Provenance: reported = stated by the company; estimated = our estimate; sample = a sample figure.

## About Axelera AI

Axelera AI is an Eindhoven startup making low-power chips that run AI models inside cameras, robots, drones and servers rather than in a distant data centre. Its 2026 round was the largest ever for an EU AI chip company.

- Raised $250M+ Series C in Feb 2026 led by Innovation Industries; over $450M raised in total.
- Europa chip (629 TOPS) shipping since Sep 2026 in Dell and Supermicro systems; 600+ customers.
- Titania data-centre chiplet, backed by up to €61.6M from EuroHPC, targeted for 2027-2028.

- Founded: 2021
- Headquarters: NL
- Status: private
- Raised: $250M
- Last round: Series C $250M, Feb 2026, led by Innovation Industries
- Website: https://axelera.ai
- People: Fabrizio Del Maffeo (Co-founder & CEO); Evangelos Eleftheriou (Co-founder & CTO)
- Partners, customers, investors: EuroHPC Joint Undertaking (partner); Samsung Catalyst Fund (investor); BlackRock (investor); Dell (partner)

Company page: https://technooptimist.io/companies/axelera-ai.md

## Related programs

Everything in Computing: https://technooptimist.io/computing.md

---

The Techno Optimist tracks every frontier of technology, program by program: https://technooptimist.io/ (a Markdown version of any page: add .md to its URL; site map for agents: https://technooptimist.io/llms.txt).
Data API (JSON, realtime stream): https://technooptimist.io/developers. People read 5 pages free, then Reader $10/month; these Markdown pages are open to agents.