EUR 850.00 For Business Accounts Only

Morningstar | The Future of Self-Driving and ADAS at Intel's Mobileye is Bright After Strong Showing at CES 2019. See Updated Analyst Note from 10 Jan 2019

At CES 2019, Intel unveiled new ninth generation desktop processors to be launched in the second quarter, 10-nanometer Sunny Cove processors to be launched for holiday 2019, and Cascade Lake server CPUs available today (for major cloud customers). Through Nervana, the firm will also be launching an inference-based NNP-I sometime in 2019 to compete against the likes of Nvidia’s T4 and Google’s TPU, though exact details weren’t disclosed. We were most interested in the firm’s Mobileye presentation, which we attended, to hear co-founder Amnon Shashua provide updates to the firm’s progress in both ADAS and self-driving. We remain positive on Intel’s future prospects and see an attractive margin of safety relative to our unchanged $65 fair value estimate for this wide-moat chip titan.

In 2018, Mobileye recorded 28 new design wins while 7 vehicle models launched with EyeQ 4 processors for advanced functionalities beyond basic ADAS. For the year, it shipped 12.4 million EyeQ chips versus 2.7 million in 2014, which represents a 46% CAGR. Highlighting the firm’s safety prowess, 12 of the 16 Euro New Car Assessment Program 2018 models with 5-star safety ratings had Mobileye inside. The latest BMW X5 has a trifocal camera powered by the EyeQ 4 for superior 3D detection and map harvesting via Mobileye’s Road Experience Management that captures mapping data in production vehicles. We continue to view Mobileye’s strategy of perfecting ADAS to bolster the necessary building blocks for autonomous vehicles as a pragmatic and economical move.

We saw multiple Nvidia autonomous driving boards at major auto supplier ZF’s booth named ZF ProAI and spanning from L2+ through L4/L5 robotaxi vehicles. ZF is also integrating Xilinx’s multiprocessor system on chip for aggregation, preprocessing, and distribution for sensor and autonomous driving data while providing low latency and high efficiency for accelerating AI performance. At Mobileye’s presentation, Shashua stressed “true redundancy” via Mobileye’s visual perception approach that seeks to achieve full autonomous-driving with cameras and then layering on top lidar and radar as true redundancies versus many current strategies that utilize sensor fusion as a crutch (integrating multiple sensor types to take a deterministic action), which we think could be flawed if one of the data inputs is compromised. Representatives of ZF agreed with our assessment, and similarly expect to see an assortment of vendors succeeding in the autonomous world beyond just Nvidia (including Xilinx and Intel via Mobileye and Altera).

We note Mobileye’s forward trifocal camera and total of 12 cameras on its vehicles have unique algorithms to extract true 3D mapping of an environment, which is considerably more cost-optimized. Finally, we saw an EyeQ 5 board for BMW 2021 series produced with auto supplier Aptiv that will be capable of L4/L5 autonomy, which bodes well for Intel/Mobileye’s self-driving prospects.

For further insight into our views on the self-driving prospects for Intel and its peers, please see our technology Observer "Automotive Chipmakers in the Fast Lane--Destination: High-Speed Growth."