EyeQ

EyeQ
Name	EyeQ
Developer	Mobileye
Type	System on a chip
Generation	Multiple
Launched	2007
Predecessor	N/A
Successor	Ongoing

EyeQ. It is a family of system-on-a-chip (SoC) microprocessors designed by Mobileye, a subsidiary of Intel, specifically for advanced driver-assistance systems (ADAS) and autonomous driving technologies. These application-specific integrated circuits (ASICs) are engineered to process data from multiple camera and sensor inputs in real-time, enabling critical functions like collision avoidance and lane keeping.

Overview

The EyeQ series represents a cornerstone of modern automotive safety and vehicle automation, powering the perception systems in millions of vehicles worldwide from manufacturers like BMW, General Motors, and Volkswagen Group. These processors are integral to achieving various levels of the SAE automation scale, from basic driver assistance to conditional autonomy. By offloading complex computer vision and deep learning tasks from general-purpose ECUs, the EyeQ architecture provides the deterministic performance required for safety-critical automotive applications.

Technology and Architecture

EyeQ chips utilize a heterogeneous computing architecture that combines multiple specialized processing cores. This includes vector processing units for image processing, programmable DSPs, and dedicated accelerators for convolutional neural network (CNN) inference, which is essential for object detection and semantic segmentation. The architecture employs an innovative memory hierarchy and interconnect fabric to manage high-bandwidth data flow from up to eight megapixel cameras simultaneously. This design philosophy emphasizes functional safety, with features aligned with the ISO 26262 standard for road vehicles.

Applications and Use Cases

Primary applications span the entire spectrum of ADAS, including adaptive cruise control, automatic emergency braking, and traffic sign recognition. In more advanced implementations, such as those deployed by Tesla in its Autopilot system or by NVIDIA-equipped prototypes, EyeQ processors enable features like highway pilot and navigate on autopilot. The technology is also pivotal for pedestrian detection, cyclist detection, and constructing a local dynamic map for robotic vehicles. Its use extends to commercial vehicles for blind spot monitoring and driver monitoring.

Development and History

Development began at Mobileye in the early 2000s, with the first-generation EyeQ1 launching in 2007 following a collaboration with STMicroelectronics. A significant milestone was the adoption of the EyeQ3 by Tesla for its first-generation Autopilot hardware suite. Subsequent generations, developed in partnership with Intel after its acquisition of the company, have seen exponential growth in computational performance. The roadmap has consistently evolved to incorporate more advanced neural network accelerators and support for sensor fusion with lidar and radar data from suppliers like Continental AG.

Performance and Specifications

Performance metrics have increased dramatically across generations, with transistor counts growing into the billions and TOPS (trillion operations per second) capabilities scaling to meet the demands of Level 2 and Level 3 automation systems. For instance, later iterations deliver vastly superior frames per second processing for high-resolution camera streams compared to early versions. Key specifications often include the number of cores, power consumption measured in watts, and support for specific automotive Ethernet protocols for high-speed in-vehicle networking.

Market and Industry Impact

The EyeQ series has profoundly shaped the automotive industry, making advanced safety features accessible and contributing to Euro NCAP five-star ratings for many vehicles. Its widespread adoption has positioned Mobileye as a dominant supplier in the ADAS market, competing with platforms from NVIDIA (Drive), Qualcomm (Snapdragon Ride), and Texas Instruments. The technology's success has accelerated industry-wide timelines for autonomous vehicle deployment and influenced regulatory discussions at bodies like the National Highway Traffic Safety Administration (NHTSA) regarding vehicle safety standards.

Category:Advanced driver-assistance systems Category:Microprocessors Category:Automotive technologies