Advanced Computing in the Age of AI | Friday, March 29, 2024

Intel-BMW and Nvidia-Mercedes Vie for Driverless Car Spotlight at CES 

(Source: Intel)

Nvidia with Mercedes-Benz and Intel with BMW made major autonomous vehicle announcements this week at the Consumer Electronics Show, with Nvidia/Mercedes vowing the release a driverless car by the end of the year and Intel/BMW saying that approximately 40 test vehicles will be on the road in the U.S. and Europe during the second half of 2017.

The Nvidia/Mercedes collaboration began three years ago, with team co-located in Sunnyvale and Stuttgart, and is part of an ongoing work focused on deep learning and artificial intelligence. Besides Mercedes-Benz, works with impressive list of other automakers, including Tesla, Volvo, Honda - and BMW.

“When our teams came together there was instant chemistry, and we share a common vision about how AI can change your driving experience, and make it more enjoyable,” NVIDIA founder and CEO Jen-Hsun Huang, who puts forward his vision for GPU-driven machine learning at every opportunity. “It is very clear that AI is going to be the future of computing. This is an endeavor that we started three years ago that we will put on the road next year – unbelievable.”

Earlier in the week Intel, in partnership with BMW Group and Mobileye, announced Intel GO, a scalable car-to-cloud system with development kits that can be used by other auto makers. The system offers a range of Intel technologies, from Atom processors to Xeon chips plus Arria 10 FPGAs.

The Intel/BMW strategy appears to rely strongly on the Intel 5G Modem, which the company said is the world’s first global 5G modem supporting both sub-6GHz bands and mmWave spectrum. It includes a compact, low-power chip kit and delivers gigabit speeds and low latency, enabling self-driving cars to make quick decisions, and transforming connectivity required for other applications, such as smart cities, drones and virtual reality applications.

On Tuesday, Intel announced plans to purchase a 15 percent ownership stake in HERE, a global provider of digital maps and location-based services. The companies will also collaborate on the research and development of a scalable proof-of-concept architecture that supports real-time updates of high-definition maps for highly and fully automated driving.

A BMW engineer in an autonomous test car. Source: BMW Group

Doug Davis, senior VP and GM of Intel’s Automated Driving Group, said the GO solution is designed to meet the thermal and safety requirements of the automotive industry while handling sensor fusion, driving policy, environment modeling, path planning, decision making and other functions. The system also offers Intel SSDs to the Intel Nervana platform for a machine and deep learning training and simulation infrastructure.

The ultimate goal, according to Klaus Fröhlich, a member of the board of management of BMW AG for Development, is production of the BMW iNEXT, BMW Group’s autonomous vehicle scheduled for delivery in 2021.

“Intel really surprised everyone with their BWM win and GO is the next step to create a broad ecosystem,” Patrick Moorhead, president of Moor Insights and Strategy, told EnterpriseTech. “With GO, Intel places itself into an exclusive group of car chip providers with NVIDIA and Qualcomm-NXP.”

Under the structure of the partnership, BMW is responsible for driving control and dynamics, safety evaluation, component integration, production of prototypes and, eventually, scaling the platform via deployment partners.

Mobileye contributes its EyeQ5 computer vision processor for interpretation of input from vision sensors, as well as localization. In combination with Intel CPU and FPGA technologies, it will be the central computing platform of the BMW iNEXT.

Mobileye will also develop with BMW the “sensor fusion solution,” creating a model of the environment surrounding the vehicle using input from vision, radar, and lidar sensors. This will establish a driving policy, including Mobileye’s reinforcement learning AI algorithms for negotiating driving situations.

According to a blog post by Davis, Intel has 49 car wins, with more than 30 Intel-based vehicle models on the road today. He also said Intel is in hundreds of autonomous test cars. He said that besides BMW, Delphi and Baidu have announced plans to use Intel technology in their autonomous vehicles.

He emphasized the significance of 5G to the evolution of self-driving. “Automated vehicles will both generate and take in huge amounts of data in order to navigate and react to sudden changes,” Davis said. “Today’s communications systems simply were not designed to handle the massive bandwidth required to support this. That’s where 5G comes in, delivering faster speeds, ultra-low latency and vehicle-to-vehicle connectivity….”

“From an industry perspective, we are already seeing savings and speed in development by sharing development costs and in pooling resources to develop a complete autonomous platform,” said Intel CEO Brian Krzanich. “The car-to-cloud system will perform with consistent, predictable behavior and is validated to the highest level of safety. That’s why this partnership is breaking new ground. We have established a dedicated team with clear, shared goals and a culture of innovation and agility and accountability.”

EnterpriseAI