Why HPC Matters: Autonomous Vehicles

High performance computing is an essential key to the self-driving vehicle — from city streets to remote mining sites.

2 whyhpcmatters autonomousvehicles bigger
Dell EMC

As they move from corporate R&D labs to public streets and highways, self-driving vehicles are getting a lot of attention. Industry observers love to talk about the potential of these futuristic cars and trucks to transform the way we move people, products and materials from one place to another. But there’s one thing some commentators may be missing when they talk about autonomous vehicles (AVs) being the next big thing. AVs have actually been with us for decades, operating efficiently in many industries.

Just ask the folks at Caterpillar, the company that introduced the industrial world to autonomous vehicles in 1996.[1}  That legacy continues today in the Caterpillar autonomous mining program, which is working to take the driver out of the company’s Cat® 793F mining trucks.

With a 240-ton payload capacity, Cat® 793F mining trucks are built for the demands of hauling copper, coal, gold, iron ore and overburden on rugged mining sites in remote areas — with or without an operator in the cab. The trucks are guided by the Cat MineStar™ System, which functions as the brains of the autonomous system. Cat MineStar software manages how each truck goes to each loader, where it takes the material to, and how it avoids other trucks on the site.

And this is where high performance computing (HPC) enters the picture. Caterpillar’s autonomous mining program generates an enormous amount of data and the corresponding need for HPC systems to mine through all that data. For this piece of the autonomous mining puzzle, Caterpillar leverages an Intel-based HPC system from Dell EMC to accelerate time to insights about product issues and optimization opportunities.

“Once you take the people out of the system, you still have to keep track of any issues that might arise on the trucks,” says Thomas McCauley, an engineering manager with the Caterpillar autonomous mining program. “Each truck has its own log, and we download that back to our main database. Then we use HPC to mine through that data to make sure everything is going well and as designed. And if it isn’t, HPC gives us the opportunity to dig into that data to find out what could be optimized, so we can address it faster.”[2]

A world with millions of AVs

Now take this big-data story and multiply it times millions of new autonomous vehicles that are coming our way soon. By 2035, more than 12 million fully autonomous vehicles are expected to be sold per year, and autos with AV features are expected to make up 25 percent of the global market, according to a study conducted by the Boston Consulting Group (BCG) in collaboration with the World Economic Forum.[3]  Each of these AVs will produce a flood of data that will need to be processed and analyzed.

A single self-driving car, for example, will generate an average of 4 terabytes of data per day, according to Intel, which is investing heavily in the enabling technologies for AVs.[4]  To process this data deluge in real time, autonomous vehicles require HPC systems on board to enable the vehicle to make spilt-second decisions, as well as in connected cloud data centers that run analytics on the constant stream of bits and bytes.

“Autonomous vehicles are the equivalent of supercomputers rolling down the highway, generating and transmitting a mind-boggling volume of data,” notes Data Center Frontier founder Rich Miller in a blog post on the data generated by AVs.[5]

“Data is truly the new currency of the automotive world,” writes Intel chief executive officer Brian Krzanich in an online editorial. “It’s not enough just to capture the data. We have to turn the data into an actionable set of insights to get the full value out of it. To do that requires an end-to-end computing solution from the car through the network and to the cloud — and strong connectivity.”

To make a big-data problem even bigger, consider this: It’s not just cars that we are talking about. Autonomous vehicles are on the job today in a wide range of commercial settings, from manufacturing plants and industrial sites to warehouses and distribution centers, and soon we will have fleets of self-driving trucks on our highways. Each of these vehicles generates a huge amount of data — and the corresponding need for HPC systems.

Democratizing HPC

There’s an important backstory here. As AVs become ubiquitous, they are going to require the data processing power of ubiquitous HPC systems. And this is yet another reason why HPC systems are heading into the mainstream. HPC systems now need to be everywhere — from the data centers that support AVs to the mining sites and manufacturing floors of the industrial world, and from the trading desks of financial service companies to the design shops of digital content creators. We’re talking about the democratization of HPC.

To enable this new era in which HPC is everywhere, technology companies are working to deliver mainstream HPC solutions that are easy to order, easy to deploy and easy to manage. These offerings, which include preconfigured and bundled solutions, help all sizes of organizations move beyond the historical barriers to HPC deployments — including the complexity of designing, configuring and ordering systems and the lack of off-the-shelf systems designed and tuned for HPC applications.

Take, for example, the Dell EMC Ready Bundle for HPC. It helps organizations meet their HPC needs with domain-specific designs that are built on tested and validated industry-standard building blocks, and all backed by a single point of contact for services and support. That’s a far cry from the complex HPC systems of the past, which required armies of engineers for system design and configuration, along with specialists for ongoing management and support.

Ultimately, while there are many unknowns about the transition to a world filled with autonomous vehicles, there is one thing we can be sure of: We are entering an era in which artificial intelligence (AI) will drive our vehicles, and HPC will drive the AI.

For a closer look at Caterpillar’s autonomous mining program, read the case study “Cat® trucks get the job done with the help of an HPC cluster from Dell EMC.”


Making a difference with HPC

High performance computing touches virtually every aspect of our lives. HPC is making weather forecasts more accurate, cancer therapies more precise, fraud protection more foolproof and products more efficient. In this series of articles, we explore these and other use cases that capitalize on HPC and its convergence with data analytics to illustrate why HPC matters to all of us.


[1]Caterpillar Facts Sheet Q4 2016.
[2]Dell EMC case study, “Autonomous Mining: Cat® trucks get the job done with the help of an HPC cluster from Dell EMC”, August 2017.
[3]Boston Consulting Group (BCG), “Autonomous Vehicle Adoption Study, January 2015.”
[4]Intel CEO Brian Krzanich, “Data is the New Oil in the Future of Automated Driving”, November 15, 2016.
[5]Rich Miller, Data Center Frontier, Autonomous Cars Could Drive a Deluge of Data Center Demand, May 24, 2017.