During a recent Download Day fireside conversation held by American AI pharmaceutical company Recursion, Nvidia CEO Huang Renxun shared his insights on similarities between drug discovery and early development in the chip industry, and discussed how Nvidia entered the healthcare industry and how it collaborated with external companies.
Nvidia is a shareholder of Recursion. Last July, Nvidia invested $50 million in Recursion to collaborate on optimizing Nvidia's cloud services and training Nvidia's AI models in the biopharmaceutical field. This investment has attracted much attention, and on the day of the news, Recursion's stock price surged by 115.29%. Nvidia is currently investing in multiple fields, with 14 outward investments in 2023. Through investment, it has entered the fields of basic large models, biopharmaceuticals, chip optical connectivity, industrial automation, and more.
"My career began 41 years ago, when computer-aided design began to make progress. Although some people had mentioned computer-aided design before, it was not until then that the combination of algorithms, fast computers, and knowledge truly introduced 'methodology' into modern chip design. The three core elements of computing power, algorithms, and professional knowledge are now flourishing in your industry (biopharmaceuticals)." When discussing the similarities between the development of the silicon chip industry and the biopharmaceutical field, Huang Renxun said.
Huang Renxun stated that chips have design rules and methods, and the challenges in the medical and health field are even more difficult because researchers must learn the behavior of organisms, understand their meaning, behavior, and characteristics. The good news is that with technology now available, researchers may only need to "click a mouse" to understand the meaning of life, thanks to data processing, machine learning, and other technologies in robotics laboratories, as well as supercomputers co built with Nvidia.
When talking about Nvidia's strategy of expanding into multiple fields, Huang Renxun stated that in addition to collaborating with Recursion, another option for Nvidia's investment is to do what others have done and do better. Pursuing fast investment returns is the norm in the business world, but this is not Nvidia's goal.
"We hope to do things that others have not done before, and things that we do not do and others will not do. Choosing this path can make a real contribution, which is our DNA." Huang Renxun said that understanding the language of life is difficult, but we can work hard. Among the three elements he talked about (algorithms, computing, and professional knowledge), NVIDIA can contribute the first two elements. Considering that NVIDIA does not have and does not aspire to have professional knowledge, NVIDIA is more willing to become a partner of others. In many industry sectors, Nvidia does not want to become a leader, but rather hopes that there will be leaders in the industry, while Nvidia hopes to focus on its unique contributions.
At last week's shareholder meeting, investors were also quite concerned about Nvidia's layout in the biopharmaceutical field. Huang Renxun stated at the time that healthcare, automotive, and digital manufacturing were the largest vertical industries served by Nvidia. Artificial intelligence has ushered in a new industrial revolution that will transform into innovation in the healthcare industry. Computers can represent, understand, and describe drugs and chemistry, enhancing people's ability to discover drugs.
In this fireside conversation, Huang Renxun also talked about Nvidia's choice of development direction. "Nvidia has not attempted to replace general-purpose computing and believes that the existence of CPUs (central processing units) is justified. Nvidia has also designed CPUs. So why not make enhancements on this basis?" Huang Renxun said. Nvidia believes that if a way can be found to make it the essence of computers and design a programming model that can handle the corresponding complexity, it can truly change the future of computing. From the beginning, Nvidia firmly believed in this. Nvidia discovered computer graphics as its first application, followed by image processing, molecular dynamics, particle physics, fluid science, etc., until one day it discovered deep learning.
Huang Renxun has repeatedly mentioned the statement that "we are only 30 days away from bankruptcy". In this fireside conversation, Huang Renxun mentioned more details. He said that Nvidia is a very confident company, on the one hand, it needs to be confident that it can do things that others cannot do, and on the other hand, it needs to have a sense of urgency. Every morning when he wakes up, he first thinks of unresolved problems that are challenging. If he doesn't face these things directly, things may get worse. He had this feeling in 1993 and still has the same consciousness today.
"A long time ago, I often shared financial information with employees and reviewed our financial statements with them every Friday. When someone pointed out that our cash flow might turn negative in June, I said, 'If we can't make money before that and can't increase revenue in the next 30 days, the company will go bankrupt.'" Huang Renxun said, "At that time, Nvidia had two options: making money and raising funds. He was not sure if Nvidia was good at either of them, so he thought the company might go bankrupt or have a turning point.". Even at a meeting after Nvidia went public, when someone mentioned rumors of the company going bankrupt within 30 days, he responded that if the company didn't give it all in the next 30 days, it could indeed go bankrupt one day in the future.
"For startups, every day is important, and the decisions, efforts, hiring, achievements, and shaping of the company culture made over the next 30 days will determine whether the company's fate is full of vitality, leading to success or bankruptcy," said Huang Renxun.