QUALCOMM ANNOUNCES EXPANSION OF SCALABLE SNAPDRAGON RIDE PLATFORM PORTFOLIO
>> Qualcomm Technologies, Inc., today announced at the company’s virtual Automotive Redefined: Technology Showcase, the expansion of its Qualcomm® Snapdragon Ride™ Platform portfolio
Qualcomm Technologies, Inc., today announced at the company’s virtual Automotive Redefined: Technology Showcase, the expansion of its Qualcomm® Snapdragon Ride™ Platform portfolio. With new safety-grade system-on-chips (SoCs) designed for automotive safety integrity level D (ASIL-D) systems, this provides the flexibility of using a single SoC for New Car Assessment Program (NCAP) Level 1 advanced driving assistance systems (ADAS) systems and Level 2 automation systems, as well as enhancing the performance of combined ADAS SoCs and artificial intelligence (AI) accelerators for up to a Level 4 automated system. The new SoCs join Snapdragon Ride Platform’s growing portfolio of comprehensive solutions designed for ADAS and autonomous driving compute systems.
These SoCs not only address the increasing complexity and computational demands of the automotive industry’s evolving zonal electronic/electrical (E/E) architecture, but they also provide a homogenous solution that serves as a central hub for high-performance compute, computer vision, artificial intelligence (AI) and multi-sensor processing with the highest level of automotive safety standards—all while harmonizing functionality with customized software implementations to address the compute, performance and safety needs for that zone or domain. The SoCs’ integrated Qualcomm® Car-to-Cloud Soft-SKU capabilities offer automakers the flexibility to upgrade features and performance capabilities, even after the vehicle has left the dealership, with over-the-air (OTA) updates, which allows the end-consumer to continuously access the latest updates and features throughout the life-cycle of the vehicle.
With the latest SoC additions to the roadmap, Snapdragon Ride supports the spectrum of ADAS/AD functionality— ranging from NCAP windshield-mount ADAS solutions (Level 1), to active safety with conditional automation (Level 2/2+), to fully autonomous driving systems (Level 4). The Snapdragon Ride platform portfolio is comprised of one of the automotive industry’s most advanced, scalable, and highly customizable automated driving SoC platforms built on 5nm process technology, engineered to not only support automakers and Tier-1s to produce more power-efficient and thermal designs, but to also provide alternatives for customization opportunities, while also charting a course for next-generation automotive architecture evolution. Snapdragon Ride now scales from 10 Tera Operations Per Second (TOPS) for windshield-mount ADAS cameras in a sub-5W power budget to over 700 TOPS for fully automated driving solutions. Since its introduction in 2020, Snapdragon Ride has gained momentum with global automakers and Tier-1 suppliers who are currently developing their next generation fleets with Snapdragon Ride.
Snapdragon Ride offers an open programmable framework that allows automakers and Tier-1 suppliers to customize the platform for their unique requirements for camera perception, sensor fusion, drive policy, automated parking, driver monitoring and more. To accelerate the deployment of cutting-edge ADAS/AD solutions, Snapdragon Ride continues to expand its software ecosystem with various industry-proven and leading stacks for vision perception, parking and driver monitoring. These SoCs are expected to provide global automakers and Tier-1 suppliers with an opportunity to select one or a combination of these various stacks based on Snapdragon Ride.
> Arriver™: Veoneer has created its software brand, Arriver, which represents Veoneer’s European NCAP award-winning perception and driving policy software stack. Qualcomm Technologies plans to offer SoCs pre-integrated and pre-validated with Arriver vision perception and driving policy software stack aimed to address the growing complexities associated with various tiers of solutions from NCAP to L2+ ADAS, and beyond.
> Valeo’s Park4U®: Qualcomm Technologies is working with Valeo, a leader in ADAS and parking automation, to support Valeo’s latest generation of Park4U® on the Snapdragon Ride Platform. Park4U® is a scalable software platform that uses multiple cameras, ultrasonic sensors as well as optional radars or lidars to enable automatic parking functions up to SAE Level 4. Snapdragon Ride offers the computing power to run Valeo’s software stack in combination with other vehicle functions.
> Seeing Machines e-DME: broadening the capability of Snapdragon Ride to deliver a comprehensive solution for higher levels of automation, Qualcomm Technologies is working with Seeing Machines to design a leading DMS solution available to automakers and Tier-1 suppliers. This latest DMS technology brings Seeing Machines’ optical and imaging systems expertise, as well as deep acceleration solutions to Snapdragon Ride. The flexible hardware system extension, utilizing the high-performance compute capabilities of Snapdragon Ride, allows single or multi-camera driver and in-cabin monitoring systems development.
Snapdragon Ride is designed as an open and programmable platform, comprised of a scalable software framework with multiple safety-ready Board Support Package (BSP) configurations that support:
> Safety Real-Time Operating Systems (RTOS): Blackberry® QNX® OS for Safety; Safe RT Linux; WITTENSTEIN high integrity systems SAFERTOS®
> Hypervisors: Blackberry® QNX® Hypervisor for Safety
> Classic and Adaptive Automotive Open System Architecture (AUTOSAR) solutions
> Open and programmable safety software development kit (SDK) with comprehensive safety solutions for sensor pipeline; provides the ability to support combinations of virtualization and containerization for optimal memory, performance and functional safety and isolation needs in mixed criticality architectures, allowing multi-domain integration on a single zonal compute platform
> Extensive set of pre-built safety libraries for imaging, computer vision and standard arithmetic functions optimized to specific hardware architectures
> AI tools for improving model efficiencies, as well as optimizing run-time on heterogeneous compute units
> Middleware support for integration of multiple software stacks (driving, parking, driver monitoring and AR-HMI) developed by Qualcomm Technologies and 3rd party companies
> Comprehensive tools for smart data collection and automated annotation during development and post-deployment phase; Snapdragon Ride provides multiple tools using AI technologies to have smart triggers during data collection, as well as automated ground truth annotation based on cross sensor calibration
> Localization: Qualcomm Technologies offers an optimized and lightweight Qualcomm® Vision Enhanced Precise Positioning (VEPP) engine that allows ADAS systems to localize in a lane with the dependency only on simple features detected on the camera, fused with GNSS and other odometry sensors.
Snapdragon Ride-based SoC and Accelerator L2+ to L4 platforms are now available and are expected to be available in production vehicles beginning in 2022. Snapdragon Ride SoCs for NCAP to L2+, with integrated stacks, are planned for 2024 production programs.
“The Snapdragon Ride Platform portfolio uniquely accelerates the deployment of ADAS and autonomous technology to automakers that can be customized and scaled across vehicle segments,” said Nakul Duggal, senior vice president & GM, automotive, Qualcomm Technologies Inc. “Our latest expansion of the Snapdragon Ride Platform portfolio and the integration of core ADAS/AD stacks for vision perception, drive policy, parking and driver monitoring holistically address the increasing complexity and computational demands that automakers are navigating today while also delivering automotive safety and functionality for drivers and passengers.”