Offers on-device supports for LLMs and VLMs
Israeli chip startup Hailo has released its second-generation AI chip, Hailo-10H.
The chips offer on-device capabilities, including support for large-language models (LLMs) and vision-language models (VLMs), allowing customers to run generative AI applications locally without using cloud-based generative AI services.
Built using “proprietary, powerful, and scalable structure-driven dataflow architecture that takes advantage of the core properties of neural networks,” the company said Hailo-H10 is optimized for Edge environments and features 40 tops of INT4 performance, with an average power consumption of 2.5W.
“With the Hailo-10H now available for order, we’re taking another major step toward our mission of making AI accessible to all,” said Orr Danon, CEO and co-founder of Hailo. “This is the first discrete AI processor to bring real generative AI performance to the Edge, combining high efficiency, cost-effectiveness, and a robust software ecosystem.”
Founded by Danon and Avi Baum in 2017, Hailo designs chips to run AI workloads on Edge devices, targeting sectors such as personal compute, automotive, security, and retail.
In April 2024, the company closed an extended Series C funding round that saw it raise $120 million from investors including OurCrowd, Talcar, Comasco, and Poalim Equity.
To date, the company has raised $564m over nine rounds.
