Eyeq4 Datasheet [ 720p ]

Note: The datasheet explicitly states that third-party CUDA or OpenCL code is not supported.

The EyeQ4 is frequently chosen by major automotive Tier-1 suppliers and OEMs for its safety-critical capabilities.

The EyeQ4 is offered in a package.

Traffic lights, construction barriers, regulatory signs, and lane demarcations (solid, dashed, colored).

: Used by major manufacturers including BMW, Nissan, Honda, and General Motors. eyeq4 datasheet

💡 : For specific implementation, designers often pair the EyeQ4 with a dedicated power management unit like the TI LP875761-Q1 to handle the SoC's core rail requirements.

When integrating computational hardware into vehicle camera housings (such as the ZF S-Cam4 mounted behind the rearview mirror), form factor is highly critical.

For engineers reading an , the following electrical and physical parameters are critical:

| Specification | Details | | :--- | :--- | | | 28nm FD-SOI (Fully Depleted Silicon On Insulator) by STMicroelectronics | | Total Compute Cores | 14 cores in total (4 CPU cores + 10 specialized accelerators) | | CPU Cores | 32-bit MIPS InterAptiv cores (5 total; 4x I-class, 1x M-class) clocked at ~1 GHz | | Vector Accelerators | 6x Vector Microcode Processors (VMP) : For mature, efficient vision tasks, proven in EyeQ2 and EyeQ3. | | Novel Accelerators | 2x Multithreaded Processing Clusters (MPC) : Versatile accelerators for complex vision algorithms, more energy-efficient than a GPU for specific tasks. 2x Programmable Macro Arrays (PMA) : Delivers compute density near that of fixed-function hardware but with full programmability. | | Memory Support | LPDDR4 memory; detailed specifications for power rails (e.g., DDR 1.1V) are available in official design guidelines. | Note: The datasheet explicitly states that third-party CUDA

Because the EyeQ4 is used in mission-critical automotive systems, it is built to survive harsh environmental and operational stressors. The hardware conforms to strict automotive industry standards, including AEC-Q100 qualification for reliability. Furthermore, the architecture is designed with functional safety in mind, supporting ASIL (Automotive Safety Integrity Level) standards required by global regulators to prevent critical system failures. Exploring the EyeQ Ecosystem

: The most capable version, supporting sensor fusion with radars and scanning-beam lasers. Applications & Industry Use

Utilizing a 4-camera setup for 360-degree monitoring.

Features four CPU cores capable of executing four hardware threads each (totaling 16 threads). These cores handle upper-level system tasks, application scheduling, interface buses, and general driving policy tracking. Vision and Deep Learning Accelerators combined with ZF’s mid-range radar

The EyeQ4, combined with ZF’s mid-range radar, is used to optimize lateral and longitudinal control in Toyota vehicles. 5. Conclusion: The Legacy of EyeQ4

Four CPU cores, each featuring four hardware threads. Performance Metrics

Mobileye developed multiple versions of the chip to support different vehicle capabilities: EyeQ4-High (EyeQ4H)