Empower embedded edge applications with state-of-the-art AI inference capabilities.

Aetina single board computer with Rockchip RK3588
Supported by Aetina RK3588 Industrial Motherboard designed for diverse applications where energy efficient Edge AI in a small form factor is critical.
Everything you need to get started in a single box pre-integrated Metis M.2 card and pre-installed Voyager SDK.
A simplified Edge AI inference - run a preloaded neural network in under ten minutes and leverage industry-standard API combined with built-in evaluation tools.
A growing Model Zoo with a broad coverage of computer vision networks for object detection, image classification, pose estimation and segmentation.
| Metis | 1x Metis M.2 card with 1 GB memory |
| CPU | 8-core 64-bit ARM (4x Cortex-A76 and 4x Cortex-A55), frequency up to 2.4 GHz |
| GPU | ARM Mali-G610 MP4 quad-core GPU |
| Memory | Onboard LPDDR4 8 GB |
| Storage | 32 GB eMMC |
| Operating System | Linux: Ubuntu Desktop 20.04 with Voyager SDK in a Docker platform |
| Ethernet/USB/COM | 1x HDMI, 2x GbE, 2x USB, 2x M.2 Key |
| Power | DC-in 12 V / DC Jack 4-pin |
| Temperature | 0°C~70°C |
Metis AI Accelerator Cards are compatible with any host system that satisfies the requirements outlined in the datasheets. The systems below are validated Aetina examples from Axelera and its partners. If you are interested in any of these systems, please contact us by filling out the form below.
| Aetina Host Systems | ||
| Manufacturer | Host System | Description |
| Metis 4-chip PCIe | ||
| Aetina | AIP-FR68 | Intel Core i7/i9, desktop AI workstation |
| Metis M.2 Rev 1.1 | ||
| Aetina | Mure 7-A1 | Intel Core i3/i5/i7 (13th Gen), edge AI server |