As we pass the midpoint of 2022, the possibility of seeing autonomous, driverless vehicles at scale throughout major cities in the United States seems more attainable than ever. I have covered self-driving vehicles for years now, and it’s not an easy puzzle to solve. There are many considerations to balance, including compute performance, power consumption, cost, and the thermal capacities of the vehicle. Few companies are willing to take an honest stab at powering software-defined AVs, but Arm is one of them. Earlier this year, I wrote about the importance of the Arm image signal processor (ISP) to advanced driver assist systems (ADAS). You can read the piece here.
Since autonomous vehicles are starting to reach critical mass, vehicle and chip manufacturers must collaborate very closely on the hardware and software it takes to power them. Late last month, Arm announced its partnership with robotaxi company Cruise to power current and next-generation vehicles. Although Cruise is just starting to monetize its robotaxi services in San Francisco, the company has much promise. Let's dig into the details of Arm's new partnership with Cruise.
Arm and Cruise partnership
Late last month, Arm announced that Cruise would leverage Arm CPUs and IP to further advance driverless vehicles. According to Dipti Vachani, Arm SVP and GM, Automotive and IoT Line of Business, Cruise and Arm have been working together privately, behind the scenes for years now on solutions for driverless vehicles. Only now is the partnership coming to light. For those that aren't familiar with the company, Cruise is the first autonomous vehicle company to commercialize a robotaxi fleet in a major U.S. city. For now, the company is deploying a robotaxi fleet in San Francisco with the hope of expanding to other major U.S. cities in the future. I think the effort will be worthwhile as I can think of dozens of applications for driverless vehicles. Any situation where parking is a nightmare would be excellent for a robotaxi, and every major city seems to have parking issues.
Powering the future of autonomous vehicles is right in Arm’s wheelhouse; Arm is excellent at developing high-performance CPUs that consume minimal power. It’s clear to me why Cruise would leverage Arm technology as it scales out its autonomous platforms, as well as look to devices from across the Arm ecosystem to help on their future autonomous platforms. There are a couple ways that Cruise will put Arm technology to use including taking advantage of Arm's automotive-enhanced (AE) line of IP focused on compute efficiency and leveraging Arm’s high performance CPUs to improve the efficiency of Cruise’s compute platform. It takes the right software and hardware to make this equation work properly, and it needs to work because safety is priority number one. The competency of a vehicle's computing system can be the difference between life and death in some circumstances, but Arm seems up to the task. Arm has a 30-year track record of excellent performance per watt in mobile and IoT applications. The company also has all major cloud providers using Arm with the recent announcement of Google Cloud using Arm Neoverse in its Tau T2A instances. The customer support for Arm computing and software is there.
This partnership between Arm and Cruise makes sense because it allows Arm to keep doing what its good at: high-performance, low-power compute, while enabling an innovator like Cruise to continue to innovate and produce an AV service for the public. The fruit of these partnerships takes time to see in the market, but since Arm has been working behind the scenes with Cruise for years, we can enjoy them much quicker. Further to that, Arm is also working with leading partners in the industry on a standardized framework called SOAFEE. According to Arm, SOAFEE is a standardized framework "that enhances proven cloud-native technologies with the real-time and safety features required in automotive applications." There is plenty of optimizations and work to be still done to reach mass deployment of autonomous vehicles, but Cruise is headed in the right direction with the help of Arm.
Cruise driverless robotaxi demo video
One of the more exciting parts of this announcement was the video demo of a Cruise AV operating a fully autonomous, electric, and a driverless vehicle in San Francisco, CA. In the video, Rene Haas, Arm CEO, and Kyle Vogt, Cruise CEO, take a ride in a Cruise robotaxi named "Lindyhop" as it operates driverless through the streets of San Francisco. You can check out the full video here. Over the last few years, I have ridden in some driverless vehicles myself, and I have to say, it is a nerve-racking experience at first. To watch a car with no driver drive you from place to place makes you appreciate the undertaking it is to develop and validate a self-driving car.
It's promising to watch demos and to see auto manufacturers and chip makers working together on hardware and software to produce a functioning driverless vehicle. Monetizing Cruise's effort so far is another hurdle as creating new technology like AVs is a capital-intensive task. Cruise is still confident in its long-term revenue potential and is targeting $50 billion a year in revenue from automated vehicle services and technology by 2030. I believe in the concept and technology; it is still to be determined if the company can execute its long-term strategy.
From a technology perspective, Cruise is making the right bet by working with an experienced SoC partner like Arm. Partnerships like this give me confidence that fully autonomous vehicles in mass deployment are closer than we think. Deep collaboration between chip and auto vehicle manufacturers will be the driving factor in getting driverless vehicles to mass production and adoption. I think we will see more of these partnerships between Arm and AV manufacturers materialize soon because of Arm’s proven track record of outstanding performance per watt. It's good to see Arm, Qualcomm, and Intel in competition in the AV space as well because it will fuel the innovation needed to get to critical mass.
There is an undeniable need for safe, driverless AVs. What’s really interesting is that Arm doesn’t actually design or manufacture end parts, but IP. Cruise is working at the very top of the value chain with Arm. Continued execution seems to be the most important in the coming months as the companies continue to develop hardware and software to create the next generation of Cruise robotaxis. There is plenty of more work to be done, but for Cruise and Arm, I think both companies will benefit from this partnership as we usher into a new era of autonomous vehicles.