Trusted By
One Platform for Edge AI
Any edge application, any network, any model, any modality, any framework, any sensor, any resolution
Edge ML at Your Own Pace
Deploy and scale at the embedded edge.
Effortlessly Scale and Deploy at the Embedded Edge
Performance
10x performance
12x faster end-to-end pipeline FPS than PCIe ML accelerator.
Power Efficient
40%-400% better FPS/watt
Significant advantage over competition in both performance and power.