eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
While it is Thanksgiving week here in the U.S., in China it is business as usual, and this is also the week for Nvidia’s GTC Conference (last year was impressive). Nvidia is all in for autonomous cars, and its Xavier platform has been gaining massive support from firms such as Continental and Veoneer, which are making it part of their own supplier-side autonomous car solutions. Volvo is using the platform to close in on Tesla; currently Volvo is the closest to Tesla in terms of applied technology.
But what makes China interesting is that it is the most aggressive in terms of both autonomous car and electric car development and deployment. Currently the U.S. leads, but it looks to many of us that China will pass the U.S. in terms of electric autonomous cars on the road before 2025—maybe even as soon as the end of 2020, when we should first get to a critical mass of level 2+ autonomous cars on the road.
Autonomous car tech has levels akin to nuclear physics. Level 2+ means it starts at Level 2, but future software upgrades may allow for Level 3 or 4 (Level 5 will likely always require a hardware change). More information on that here.
Let’s talk about Nvidia GTC China and autonomous cars this week.
The Underestimated Importance of Autonomous Electric Cars
If you step back and think about it, the autonomous car is slated to be the first AI robot at scale. Yes, it looks more like a car than Commander Data from “Star Trek: Next Generation,” but once we reach Level 5 it will be able to act fully independently and do complex tasks. Granted, these tasks mostly have to do with getting you from point A to point B but, to do this, the car has to be able to read signs, it has to be able to differentiate between a variety of objects, it has to be able to instantly analyze and respond to threats, and it has to be able to react to the ever-changing environment around it.
Now this same set of skills could be put in other form factors ranging from transportation (trucks, buses, planes, drones, etc.) to industrial equipment to ever more-versatile robots. The critical elements are the same, and they revolve around the number and effectiveness of the sensors, the intelligence of the AI and the availability of a viable power supply.
This last is being aggressively addressed with electric cars and the massive amount of investment in revolutionary battery technology that would allow higher energy density and reduce the cost, weight and space needed for these currently relatively inefficient batteries.
If you can significantly enhance all three things using personal robotics, flying cars and even electric-powered personal flying appliances (rocket belts without the rockets) become viable. A good chunk of the next industrial revolution will undoubtedly revolve around the advances initially from autonomous cars; thus, dominating this segment could pay massive dividends.
Nvidia
Nvidia saw this opportunity well over a decade ago, which is why it is in the lead with a packaged solution. It not only created the hardware and software but also a system to use machine learning and inference to more rapidly create ever smaller brains for this autonomous car future. (It is amazing what capabilities the platform contains.)
Therefore, the company is the busiest one licensing this technology or selling it outright, largely because it is simply the furthest along in this race to create a generation of diverse robots. But it shouldn’t be lost on anyone that someone likely could take the AGX Xavier platform and turn it into the most advanced digital assistant, home or business monitor, or even use it as the key component to automate a variety of tasks, once trained. And, once trained, the training can be instantly transferred to other Xavier products to instantly allow someone to significantly scale out the result.
You could have a truly smart house or office in which the solution uses intelligence to make complex decisions in your absence or when you are otherwise distracted. At the very least, it could automatically discern whether you are at risk and either suggest a remedy or automatically call for the help that is both closest and the most able to deal with your problem.
Having this leadership effectively puts Nvidia on the cutting edge of the next Industrial Revolution, Industry 4.0.
China
They key to advancing will be having as much real-world data as possible. China is moving aggressively to deploy autonomous electric cars and, thanks to a lack of regulation, is slated to get to critical mass with this technology early next decade. The data the country captures due to this rollout should provide the competitive edge to countries that aggressively deploy the technology. China is currently on that cutting edge for autonomous electric cars and, as a result, looks to be able to get to critical mass far more quickly thanks largely to massive government support and help in navigating China’s unique bureaucracy.
The result appears to be that China will get to critical mass with autonomous cars and have a significant lead with robots in general by mid-next decade. This should provide a foundation for it to expand in robotics and other ever-broader AI ventures; it sets them up to be dominant with robotics in general.
Once there, catching China or Nvidia from behind will likely be impossible.
Wrapping Up
Nvidia’s GTC conference in China is where many thought leaders are this week looking at the future, not just of autonomous cars, but of autonomous everything, particularly robotics. Now, there are some growing concerns about this technology being used for “killer robots,” which will need to properly be addressed. However, both Nvidia and China are moving toward market dominance with applied AI and robotics, and at GTC China this was extremely self-evident.
In the end, at events like GTC China we are seeing the future and we are likely seeing the early indications of who will dominate it.
Rob Enderle is a principal at Enderle Group. He is an award-winning analyst and a longtime contributor to QuinStreet publications and Pund-IT.