AI & ML


Photonic chip to slash AI energy consumption

29 August 2025 AI & ML

Arago, a Paris and Silicon Valley-based deeptech startup pioneering a new class of energy-efficient AI chips powered by light, has raised $26 million in seed funding to accelerate the commercialisation of its photonic processor, codenamed ‘JEF’.

The surge in AI adoption has led to exponential – and unsustainable – growth in energy consumption. Arago’s mission is to radically rethink AI’s compute infrastructure so that the technology can have a viable future for everyone.

A new era of photon-powered AI

The heart of ‘JEF’ is a proprietary photonic technology; instead of transistors, it uses lasers to process data with photons – particles of light that generate far less heat than electrons in conventional processors. By harnessing these inherent advantages, the chip is designed to deliver 10x lower energy consumption than today’s leading GPUs at equivalent performance and cost.

Early results demonstrate that ‘JEF’ can run AI models from industry-standard software frameworks, while staying fully compatible with the existing AI ecosystem, compute infrastructure, and manufacturing processes. It also sidesteps the technical barriers that have historically limited the performance of photonic and GPU-alternative processors.

Nicolas Muller, Arago’s CEO and co-founder, said, “To build a product that is not only high-performing, but also truly usable, it is critical to deeply understand the constraints of integrating a component based on a different compute principle into the broader ecosystem. We do not have the luxury of waiting for the ecosystem to adapt – our technology needs to be compatible with everything from manufacturing processes to the AI software stack from day one.”

“Innovation in computing is one of the hardest deeptech challenges,” added Thong Le Hoang, co-founder of Visionaries Tomorrow. “Arago’s superpower lies in its first-principles vision of the future, combined with a deep understanding across digital, analogue, and photonic approaches. The result is a unique hybrid architecture built on the latest industry breakthroughs and engineered to deliver step-change performance gains in the near term.”

Founded less than a year ago by Nicolas Muller, Eliott Sarrey and Ambroise Müller – whose combined expertise spans photonics, electronics, software, mathematics and machine learning – Arago has assembled a lean team of 20 experts from the world’s leading tech companies and research labs.

“Arago is creating a ‘DeepSeek moment’ for AI chips,” said Earlybird co-founder Hendrik Brandis. “This technology has the potential to defy the laws of AI compute, using only a fraction of the resources. The Arago team is executing towards this moment with unseen velocity and frugality.”

Arago will use its funding to accelerate product development toward commercialisation, grow its team across France, North America, and Israel, and deepen business partnerships to scale its growth.




Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

Compact direct Time-of-Flight 3D LiDAR module
Altron Arrow AI & ML DSP, Micros & Memory
The VL53L9 from STMicroelectronics is the first direct Time-of-Flight (dToF) 3D LiDAR all-in-one module in ST’s portfolio, offering a resolution of 2,3K zones, wide field of view, on-chip processing, 100 frames per second, and sensing range from 5 centimeters to 9 meters.

Read more...
SoC brings scalable edge AI to life
Altron Arrow AI & ML
The NXP i.MX 937 applications processor is optimised for performance without excessive power draw and bridges the gap between entry-level chips and high-end processors.

Read more...
Wall-mount cooling for Edge IT
AI & ML
Designed to run 24/7 in critical compute environments, Vertiv CoolPhase Wall delivers up to 60% greater airflow than comfort cooling.

Read more...
Compact direct Time-of-Flight 3D LiDAR module
Altron Arrow Opto-Electronics Electronics Technology AI & ML
The VL53L9 from STMicroelectronics is the first direct Time-of-Flight (dToF) 3D LiDAR all-in-one module in ST’s portfolio, offering a resolution of 2,3K zones, wide field of view, on-chip processing, 100 frames per second, and sensing range from 5 centimeters to 9 meters.

Read more...
Edge AI improves condition monitoring
EBV Electrolink AI & ML
STMicroelectronics has introduced the IIS3DWB10IS, an intelligent MEMS vibration sensor designed for high-performance industrial condition monitoring applications.

Read more...
Tiny SoM integrates NXP i.MX 91
AI & ML
The QS91 module from Direct Insight utilises the low-cost single core i.MX 91 to deliver secure, energy-efficient Linux capabilities for edge applications.

Read more...
Robotics platform powers intelligent systems
Vepac Electronics AI & ML
The CEXD-INTRBL open robotics development system combines high-performance AI processing, sensor integration, and motion control into a compact embedded platform.

Read more...
SensorGPT speeds intelligent IoT deployment
AI & ML
TDK has introduced SensorGPT, a new technology designed to accelerate the deployment of intelligent edge AI and smart IoT applications by dramatically reducing reliance on real-world sensor data.

Read more...
Ideal entry point into edge AI
Avnet Silica AI & ML
congatec has released new COM Express modules with Intel Core Series 3 processors for cost- and energy-optimised embedded computing applications.

Read more...
Analogue reservoir AI chip capable of real-time learning
Altron Arrow AI & ML
TDK Corporation has jointly developed a prototype of a reservoir AI chip using an analogue electronic circuit that mimics the cerebellum with Hokkaido University.

Read more...









While every effort has been made to ensure the accuracy of the information contained herein, the publisher and its agents cannot be held responsible for any errors contained, or any loss incurred as a result. Articles published do not necessarily reflect the views of the publishers. The editor reserves the right to alter or cut copy. Articles submitted are deemed to have been cleared for publication. Advertisements and company contact details are published as provided by the advertiser. Technews Publishing (Pty) Ltd cannot be held responsible for the accuracy or veracity of supplied material.




© Technews Publishing (Pty) Ltd | All Rights Reserved