Log in

You don’t have an account? Register

Products / Automotive

Redefining mobility.

Ambarella’s automotive processors are engineered for the road ahead, combining pioneering research in environmental perception with the latest advances in deep neural network (DNN) performance. Our products empower OEMs and Tier-1 suppliers to improve vehicle safety by delivering exceptional perception capability and image quality in low-light or high-contrast environments, advanced computer vision processing at 12-megapixel (12MP) with higher resolutions for superior image recognition, long- and short-range stereovision capabilities, recording functionality, low-bitrate streaming, and AEC-Q100 or ASIL qualification—all using single-chip, ultra-low-power designs.

OEM & Auto-Qualified

CV3-AD685

Ambarella’s ASIL B(D)-compliant CV3-AD685 artificial intelligence (AI) domain controller system on chip (SoC) supports multi-sensor perception, fusion and path planning for L2+ to L4 autonomous driving, and premium advanced driver assistance systems (ADAS) products.

Key Features

  • Neural vector processor (NVP) with industry-leading AI performance per watt and neural network (NN)-based processing to enable detection, classification, tracking, and more
  • General vector processor (GVP) for offloading classical computer vision and radar processing
  • Processing island targeted to meet ASIL B requirements; safety island targeted to meet ASIL D requirements
  • An ideal platform for efficiently implementing assisted and autonomous driving for L2+ to L4 vehicles, multi-sensor ADAS, DMS and in-cabin solutions, multi-channel electronic mirrors with BSD, and intelligent parking assistance and automated parking systems

CV3-AD

Ambarella’s ASIL B(D)-compliant CV3-AD domain controller system on chip (SoC) provides industry-leading AI performance per watt for neural network (NN) computation, with a performance increase of over 40x compared to Ambarella’s prior automotive family of SoCs. In addition, CV3-AD includes a general vector processor (GVP), an advanced image processor, a dense stereo and optical flow engine, up to 16 Arm® Cortex®-A78AE CPUs, and an automotive GPU, in a single SoC.

Key Features

  • Neural network based processing to enable detection, classification, tracking, and more
  • Neural vector processor with industry-leading AI performance per watt
  • General vector processor for offloading classical computer vision and radar processing, and floating-point intensive algorithms
  • Ideal platform for implementing autonomous driving for vehicles from L2+ to L4, single- and multi-camera ADAS, DMS and in-cabin solutions, single- and multi-channel electronic mirrors with BSD, and intelligent parking assistance systems.
  • Processing island targeted to meet ASIL B requirements; safety island targeted to meet ASIL D requirements

CV2FS

Designed for front-facing stereovision ADAS cameras, as well as computer vision ECUs for L2+ and higher levels of autonomy, the CV2FS chip features CVflow® AI processing and ASIL C compliance to enable safety-critical applications.

Key Features

  • ISO 26262 compliant to meet ASIL C
  • CVflow architecture with DNN support
  • Dense optical flow engine
  • Stereo disparity engine
  • Multi-exposure high dynamic range (HDR) processing and LED flicker mitigation
  • Real-time, hardware-accelerated fish-eye dewarping and lens distortion correction (LDC)
  • Advanced security features, including secure boot with TrustZone®, true random number generator (TRNG), one-time programmable memory (OTP), DRAM scrambling and virtualization

CV22FS

Featuring CVflow® AI processing and ASIL C compliance to enable safety-critical applications, the CV22FS chip targets forward-facing monocular ADAS cameras, as well as computer vision ECUs for L2+ and higher levels of autonomy.

Key features

  • ISO 26262 compliant to meet ASIL C
  • CVflow architecture with DNN support
  • Dense optical flow engine
  • Multi-exposure high dynamic range (HDR) processing and LED flicker mitigation
  • Real-time, hardware-accelerated fish-eye dewarping and lens distortion correction (LDC)
  • Advanced security features, including secure boot with TrustZone®, true random number generator (TRNG), one-time programmable memory (OTP), DRAM scrambling and virtualization

CV72AQ

Ambarella’s CV72AQ system on chip (SoC) provides 8MP90 / 4KP90 image processing, video encoding / decoding, and CVflow® computer vision processing in a single, low-power design. Fabricated in advanced 5 nm process technology, it achieves an industry-leading combination of low power and high performance in both human vision and computer vision applications.

Key Features

  • 8MP90 / 4KP90 video performance
  • Convolutional neural network (CNN) / deep neural network (DNN)-based processing: detection, classification, and more
  • CNN toolkit for easy porting of neural networks implemented in Caffe, TensorFlow, PyTorch, or ONNX frameworks
  • Target applications include single- / multi-channel electronic mirrors, multi-channel drive recorders / data loggers, multi-camera advanced driver assistance systems (ADAS), automatic parking assist (APA)

CV2AQ

Ideal for stereovision-based automotive sensing camera designs, Ambarella’s CV2AQ chip combines industry-leading image processing, 12MP30 video encoding / decoding, and CVflow® computer vision performance at full 12MP resolution enabling image recognition over long distances and with high accuracy.

Key features

  • AEC-Q100 Grade-2 qualified
  • Stereo processing to enable generic obstacle detection, terrain modeling, and more
  • CNN- / DNN-based monocular processing to enable classification, tracking, and more
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX
  • 12MP30 video performance
  • Advanced imaging for low-light conditions, LED flicker mitigation, wide dynamic range (WDR) and high dynamic range (HDR), automotive multi-channel smart auto exposure and auto white balance
  • Target applications: multi-channel drive recorder, AVM / surround view with auto parking assist (APA), single- / multi-channel electronic mirror, advanced driver-assistance systems (ADAS), driver / in-cabin monitoring system (DMS / CMS), and short-range / long-range depth perception cameras

CV20AQ

Featuring CVflow® AI processing, the CV20AQ chip targets stereo-camera advanced driver assistance systems (ADAS), multi-camera L2+ systems, multi-camera driver monitoring systems (DMS) and occupancy monitoring solutions (OMS), multi-channel electronic mirrors with blind spot detection, and intelligent parking assistance systems.

Key Features

  • CVflow architecture with DNN support
  • Multi-exposure high dynamic range (HDR) processing and LED flicker mitigation
  • Advanced security features, including secure boot with TrustZone®, true random number generator (TRNG), one-time programmable memory (OTP), DRAM scrambling and virtualization
  • Dense optical flow accelerator
  • Stereo disparity engine

CV22AQ

The CV22AQ delivers next-generation image processing, 12MP30 video encoding, and CVflow® computer vision performance in an optimized, low-power design.

Key features

  • AEC-Q100 Grade-2 qualified
  • CNN- / DNN-based processing to enable detection, classification, tracking, and more
  • Computer vision processor
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX
  • Advanced image processing: multi-exposure line-interleaved HDR, hardware dewarping engine support, electronic image stabilization (EIS), 3D motion-compensated temporal filtering (MCTF), and superior low-light processing
  • 12MP30 video performance
  • Euro NCAP features addressed
  • Target applications: multi-channel drive recorder, AVM / surround view with auto parking assist (APA), single- / multi-channel electronic mirror, advanced driver-assistance systems (ADAS), and driver / in-cabin monitoring system (DMS / CMS)

CV25AQ

With the lowest power consumption in the CVflow® family, the CV25AQ combines advanced image processing, 6MP30 video encoding, and CVflow computer vision performance.

Key features

  • AEC-Q100 Grade-2 qualified
  • CNN- / DNN-based processing to enable detection, classification, tracking, and more
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX
  • Advanced image processing: multi-exposure line-interleaved HDR, hardware dewarping engine support, electronic image stabilization (EIS), 3D motion-compensated temporal filtering (MCTF), and superior low-light processing
  • 6MP30 video performance
  • Target applications: multi-channel drive recorder, single- / dual-channel electronic mirror, and driver / in-cabin monitoring system (DMS / CMS)

CV28AQ

Ambarella’s AEC-Q100 qualified CV28AQ SoC combines image processing, 6MP30 video encoding / decoding, and CVflow® computer vision processing in a single, low-power design, making it an ideal choice to power the next generation of intelligent automotive devices.

Key Features

  • AEC-Q100 Grade-2 qualified
  • 6MP30 (180 MPixel/s) video performance
  • Image signal processing (ISP) throughput of 250 MPixel/s
  • Target applications: multi-channel drive recorder, single- / dual-channel electronic mirror, aftermarket advanced driver assistance systems (ADAS), and driver / in-cabin monitoring system (DMS / CMS)
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX

H32AQ

The H32AQ SoC combines 6MP30 video encoding, image / video processing, and a quad-core processor in a single, low-power design.

Key features

  • AEC-Q100 Grade-2 qualified
  • 6MP30 video performance
  • Flexible multi-streaming capability
  • Multi-sensor support
  • Superior image processing: multi-exposure line-interleaved HDR, hardware dewarping engine, electronic image stabilization, 3D motion-compensated noise reduction (MCTF), and exceptional low-light processing
  • Target applications: multi-channel drive recorders, single- / multi-channel electronic mirrors, and driver monitoring systems (driver distraction / drowsiness detection)

A12AQ

Engineered for electronic mirrors and automotive video recording systems, the A12AQ processor provides superior visibility even in low-light or high-contrast environments, delivering a combination of advanced image processing, high dynamic range (HDR), 3D noise filtering, smart auto exposure, and high-resolution encode performance.

Key features

  • Designed for single-channel electronic mirrors or single/dual-sensor video recording systems
  • AEC-Q100 Grade-2 qualified
  • High-performance automotive video engine
  • Superior image quality

Aftermarket & Dealer-Fit

CV72AX

Ambarella’s CV72AX system on chip (SoC) provides 8MP90 image processing, video encoding / decoding, and CVflow® computer vision processing in a single, low-power design.  CV72AX’s third generation CVflow architecture provides the DNN processing required for the next generation of intelligent cameras for automotive applications.

Key Features

  • Convolutional neural network (CNN) / deep neural network (DNN)-based processing: detection, classification, and more
  • CNN toolkit for easy porting of neural networks implemented in Caffe, TensorFlow, PyTorch, or ONNX frameworks
  • 8MP90 video performance

CV5A

Featuring 8K image processing, video encoding / decoding, and CVflow® computer vision processing, CV5A targets an array of automotive applications, including multi-channel drive recorders / data loggers and multi-camera aftermarket ADAS systems.

Key features

  • CNN / DNN-based processing: detection, classification, and more
  • CNN toolkit for easy porting of neural networks implemented in Caffe, TensorFlow, PyTorch, or ONNX frameworks
  • 8KP60 video performance
  • Advanced image processing: multi-exposure line-interleaved HDR, hardware dewarping engine, electronic image stabilization (EIS), multiple camera support, LED flicker mitigation, 3D motion-compensated temporal filtering (MCTF), and superior low-light processing

CV52A

Equipped with 6K image processing, video encoding / decoding, and CVflow® computer vision processing, the CV52A SoC is an ideal solution for the next generation of multi-channel drive recorders and multi-camera aftermarket ADAS systems.

Key Features

  • 6KP30 / 4KP90 video performance
  • CNN toolkit for easy porting of neural networks implemented in Caffe, TensorFlow, PyTorch, or ONNX frameworks
  • CNN / DNN based processing: detection, classification, and more
  • Advanced image processing, including HDR, dewarping, EIS, MCTF, and more

CV2A

The CV2A SoC includes dedicated stereovision acceleration, 4KP60 / 8MP60 video encoding, and CVflow® computer vision processing in a single, low-power design.

Key features

  • Stereo processing to enable generic obstacle detection, terrain modeling, and more
  • CNN- / DNN-based monocular processing to enable classification, tracking, and more
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX
  • 8MP60 video performance
  • Advanced image processing: multi-exposure line-interleaved HDR, hardware dewarping engine, electronic image stabilization (EIS), support for multiple cameras, 3D motion-compensated temporal filtering (MCTF), and exceptional low-light processing
  • Target applications: multi-channel drive recorder, aftermarket ADAS, driver / in-cabin monitoring system (DMS / CMS), and single- / multi-channel electronic mirror

CV22A

The CV22A chip delivers advanced image processing, 12MP30 video encoding, and CVflow® computer vision performance.

Key features

  • CNN- / DNN-based processing to enable detection, classification, tracking, and more
  • Computer vision processor
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX
  • Advanced image processing: multi-exposure line-interleaved HDR, hardware dewarping engine support, electronic image stabilization (EIS), 3D motion-compensated temporal filtering (MCTF), and superior low-light processing
  • 12MP30 video performance
  • Target applications: multi-channel drive recorder, single- / multi-channel electronic mirror, aftermarket ADAS, driver / in-cabin monitoring system (DMS / CMS)

CV25A

The CV25A chip delivers advanced image processing, 8MP30 video encoding, and CVflow® computer vision performance, allowing the implementation of computer vision algorithms such as forward collision warning, lane departure warning, driver monitoring, and more.

Key features

  • CNN- / DNN-based processing to enable detection, classification, tracking, and more
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX
  • Advanced image processing: multi-exposure line-interleaved HDR, hardware dewarping engine support, electronic image stabilization (EIS), 3D motion compensated temporal filtering (MCTF), and superior low-light processing
  • 8MP30 video performance
  • Target applications: multi-channel drive recorder, aftermarket ADAS, single- / dual-channel electronic mirror, and driver / in-cabin monitoring system (DMS / CMS)

CV28A

The CV28A SoC combines image processing, 8MP30 video encoding, and CVflow® computer vision performance.

Key features

  • CNN- / DNN-based processing to enable detection, classification, tracking, and more
  • CNN toolkit for easy porting with Caffe, PyTorch, TensorFlow, and ONNX
  • Advanced image processing: multi-exposure line-interleaved HDR, hardware dewarping engine support, electronic image stabilization (EIS), 3D motion compensated temporal filtering (MCTF), and superior low-light processing
  • 8MP30 video performance
  • Target applications: multi-channel drive recorder, aftermarket ADAS, single- / dual-channel electronic mirror, and driver / in-cabin monitoring system (DMS / CMS)

H22A

H22A delivers a highly efficient 12MP30 encoder / decoder in hardware along with an industry-leading image signal processor (ISP), providing outstanding imaging in low-light conditions along with high dynamic range (HDR) processing for maximum detail in high-contrast scenes.

Key features

  • 12MP30 H.264 / H.265 video compression
  • Flexible multi-stream capability
  • Multi-sensor support
  • Superior image processing: multi-exposure line-interleaved HDR, hardware dewarping engine, electronic image stabilization, 3D motion-compensated noise reduction (MCTF), and exceptional low-light processing
  • Target applications: multi-channel drive recorder and driver monitoring system (driver distraction / drowsiness detection)

H32

The H32 SoC combines image / video processing, 8MP30 video encoding, and a quad-core processor in a single, low-power design, making it an ideal choice for aftermarket dash cameras.

Key features

  • 8MP30 video performance
  • Superior image processing: multi-exposure line-interleaved HDR, hardware dewarping engine, electronic image stabilization, 3D motion-compensated noise reduction (MCTF), and exceptional low-light processing
  • Flexible multi-streaming capability

A12A

The A12A chip provides support for video recording through the front windshield, rear window, and inside the vehicle cabin, delivering single-channel quad HD (4MP, 1440p) at 30 fps, full 1080p HD at up to 60 fps, and dual-channel quad HD performance.

Key features

  • Full HD @ 60 fps or quad HD 4MP @ 30 fps
  • Two-channel quad HD @ 30fps
  • Electronic image stabilization (EIS)
  • Superior image quality with advanced night vision for low-light conditions, full-resolution oversampling, WDR, HDR, and more
  • ADAS application support