An inside look at NVIDIA

C-suite leaders from the world’s best companies turn to World 50 Group for pragmatic perspectives on urgent business topics and macro trends. As members begin to put their AI strategies into practice, World 50 conversations have recently covered everything related to AI—with technologists, people leaders, and policy experts joining small roundtables to field questions and offer real-time advice.

To further fuel AI inspiration and innovation, World 50 members took a trek to Santa Clara, California, to learn directly from the minds behind the global leader in artificial intelligence computing: NVIDIA . Below are takeaways from that experience, where members spoke intimately with senior leaders, learned about robotics, and toured an autonomous vehicle garage.

Software is the secret sauce.

Since 1993, NVIDIA has pivoted with revolutions in the PC, mobile, cloud, AI, and robotics spaces. Impressive for a chipmaker, right? Actually, NVIDIA considers itself an accelerated computing platform composed of more software than hardware. For example, some business groups at NVIDIA only build software computing platforms. This surprised executives in the room, one of whom voiced concern about generative AI’s low return on invested capital (ROIC).

• Advice from NVIDIA board member Aarti Shah , a veteran data scientist at Eli Lilly: Drive AI with conviction about how capital deployment fits with ESG strategy. She cited as an example novel drug development and asked, “What is that AI moment? How will it change society for good?” Kimberly Powell , vice president of health care at NVIDIA, praised Amgen for measuring the ROI of generative AI in labs. Amgen discovered that the antibody design process went from two years and a 50% success rate to three months and a 90% success rate.

What’s next for ERP: real-world simulations.

Since 2012, NVIDIA has worked on autonomous vehicles (AVs)—their first use case for NVIDIA Omniverse, or the industrial metaverse, that connects parts of a physical business in a digital twin. Simulations are the next step for companies using ERP systems to forecast based on financial, supply chain, and logistical data. The Omniverse “turns problems around atoms into software problems,” which can be solved with accelerated computing, explained Rev Lebaredian , vice president of Omniverse and simulation technology at NVIDIA. Lessons from NVIDIA’s work in AVs translate to other domains, such as health care. “We need redundancy and safety built into the chips that we’re putting into surgical robots, just like your Tesla,” said Powell, who noted that only one-third of the world has access to surgical care. As AI improves real-time sensor processing, surgical robots will get smaller, more autonomous, and capable of different surgeries—revealing the enterprise application for AI to scale scarce talent.

Generative AI unlocks a new form factor: humanoid robots. 

Today, NVIDIA boasts the largest number of humanoid species per square foot in the world. Their new AI agent can teach robots 30 complex skills, such as rapid pen spinning. Generative AI “essentially unlocks the possibilities of us having a proper robot brain,” said Lebaredian. “We’ve designed the world around us for humans, so it’s like the perfect form factor.” The largest markets for the Omniverse and humanoid robots include industrial manufacturing, supply chains, warehouses, retail, and factories, as much of this work consists of mundane tasks suited for automation. One member perspective: Do not confuse augmenting work with displacing jobs. “Jobs are not going away in the long run, if you can have a growing universe.”

CEOs build organizational loyalty with mission and culture.

With dozens of direct reports, NVIDIA CEO Jensen Huang runs a flat organization, where “the mission is the boss”—creating the conditions for people to do their life’s work. He discussed how NVIDIA hires for and cultivates the value of “learning with humility.” An example is Lebaredian, who started his career making movies and joined NVIDIA to make video games. Now, he makes robots. He shared his experience with NVIDIA’s pivot culture. When he joined 22 years ago, he was one of 1,000 employees. At his 20-year anniversary, NVIDIA had more than 20,000 employees, and 65% of the first 1,000 employees remained. Lebaredian credited Huang for training “a global neural network” that connects leaders and new workers who collaborate to accomplish a mission. “The best form of respect is to be a good listener,” said Huang. “Good listeners tend to be good learners.”  

Membership in World 50 is by invitation only. Inquire about eligibility at stayahead@world50.com.