Breakthrough acceleration that doesn’t break budgets
Enabling a green, fair, trusted and safe world.
Metis AIPU
Currently, AI technology has primarily been designed for cloud computing operations, which do not have the same cost, power and scalability constraints as edge devices. The hardware for edge applications requires an entirely new design considering specific computational performance, power, and economic conditions.
We are creating the future of Edge AI with industry-leading performance, usability, and efficiency at a fraction of the cost of existing solutions.
Meet our Metis AI Processing Unit (AIPU). Built for computer vision inference on the edge, with scalable technology to meet the future demands of AI acceleration workloads.
AI performance
The world is more interconnected now than ever before. Yet, 98% of the data being collected remains unused due to technology and processing limitations. Bridging this data gap requires a new generation of more powerful, efficient, and accessible computation hardware. By delivering class leading compute density that results up to 214 TOPs from a single Metis AI processing unit, Axelera AI solutions meet this demand, empowering AI applications at the Edge.
Usability
The potential for AI is simply massive. However, the complexity of mastering AI can be challenging, for end-users as well as system integrators and developers. By creating integrated solutions that are easy to implement and use, Axelera AI is simplifying AI at the Edge. Not just delivering raw compute performance but doing so accurately and efficiently. We are also going beyond just the accelerator technology and looking to build a fully interconnected ecosystem of support.
Our software platform, Voyager SDK, makes it easy to build high-performance AI applications.
Minimal Power
Power consumption is a critical factor when it comes to Edge applications, and for many applications will only have limited power budgets. So Axelera AI offers class leading compute density with exceptional core efficiency. The result means systems can easily crunch data without draining power or running hot, with a typical use case requiring just a few Watts. Tackling the Compute Energy vs. Global Energy Production challenge head-on.
Edge price point
Until now AI systems and applications have relied on the computational performance of large, power hungry and expensive hardware. However, fully unlocking the potential of AI, especially at the Edge, requires a dramatic increase in FPS/$ which will enable complex AI applications to run on-device at the Edge. Running the industry-standard benchmark of ResNet50v1, the Metis PCIe system can achieve a record-breaking 16.4 fps/$, which truly is breakthrough AI acceleration.
