(Yicai Global) Nov. 29 -- "If we have intelligence, we can solve any problem," Jensen Huang, co-founder and chief executive of computing graphics giant Nvidia Corp. [NASDAQ:NVDA], told Yicai Global in an interview recently, adding that artificial intelligence is likely the most important technology invented.
Huang said during a keynote speech at the Nvidia GPU Technology Conference in Beijing, "the more GPUs [graphic processing units] you buy, the more money you save," a seeming attempt to make more sales but also the truth, he added in the interview. GPU performance is so much greater than traditional CPUs (central processing units), buying one GPU could save you from buying 100 CPUs, he said, comparing the difference between the two to air and rail transport.
Nobody could have foreseen the AI boom, he continued. "People thought the first generation of AI was going to happen by the 70s. So when IBM, with its Deep Blue, was able to beat the best chess player in the world, people thought the era of AI was there, but in fact is was still so far away.
"Human intelligence is very difficult to describe. How we learn and how we generalize knowledge has been something that computer scientists and neuroscientists have been trying to understand for a long time. Then all of a sudden, this new approach came along, called deep learning."
Deep learning was a failure at first, Huang said, attributing the problems to a lack of data and computing ability. "It wasn't until only seven or eight years ago that we were able to use GPUs to train huge amounts of data," he continued, noting that this was what people refer to as the 'Big Bang' of artificial intelligence. Deep learning, new AI, copious amounts of data and having GPUs to process it all is what led to that Big Bang. "It's a very exciting time," Huang said.
When asked how Nvidia plans to stay ahead amid fierce competition in the coming year, Huang said a focus on a platform will meet industry needs, and believes the architecture the firm has developed over the past 15 years will prove a stronghold.
The company is still very concentrated on gaming, he noted, clarifying that computer gaming is still the Nvidia's core business and the move into AI is a supplement, not a change of direction. Video gaming is a huge market, and is going to become even bigger. Due to its technological complexity, it's a great driver for new technologies, Huang said.
"This is a technology that can think," he added. "If we can automate intelligence, imagine the problems we can solve. This could very well be humanity's greatest invention. That's why it's very likely that artificial intelligence is going to be a very big business for us."
Huang believes that PC gaming still has a solid future, despite the advancement of mobile gaming. Video games are a fantastic form of entertainment because they're engaging, challenging and involve solving problems. People who play video games are great problem solvers, he added.
Nvidia recently partnered up with a number of startups in pursuit of autonomous driving technology, a growing industry worldwide and particularly in China.
The company began working on autonomous driving about five or six years ago when it realized that artificial intelligence and deep learning would help it contribute to the industry.
Driving is a very complicated skill that humans can do very naturally, Huang said. It is impossible to describe exactly how people drive; there are too many situations, too many things happening all at the same time. "We realized that without intelligence, it [figuring out how humans drive] was impossible.
"Finally, about six years ago, we realized that the evolution of deep learning was going to help us solve this problem, so we decided to jump in and try to build a platform that every transportation company in the world can [use to] turn everything that moves into an autonomous vehicle. A truck, car, tricycle, airplane, boat… anything that moves we believe eventually should be autonomous, so we created this platform, we call it the Nvidia Drive.
"Nvidia drive is the world's first computer that is designed from the ground up to be completely autonomous. The processor, the operating system, all of the APIs [application programming interfaces] and middleware and all of the algorithms on top are designed for autonomous vehicles."
The company decided to keep the platform open, Huang continued, so instead of building just one car, the platform can be used by startups, major corporations, logistics firms and even taxi companies. There are already 145 startups around the world running Nvidia Drive, and about 50 original equipment manufacturing projects. "Autonomous vehicles are going to completely reshape society and our economy," he added, saying he believes the technology will make travel safer and more efficient.
"We can completely change the shape of cities, reduce the cost of goods, protect lives, reduce insurance and emergency medical costs," he said. "Hospitals will change, insurance companies will change, our roads will change, parking lots will change, parks will change, and everything will change. I'm so excited about this one vertical market, because it's just hugely important to the future."
Huang is confident about Nvidia's future in video games, despite increased competition from rival Intel Corp. [NYSE:INTC]; and believes the company will also thrive in autonomous driving and deep learning. "Every technology company should try to develop a strategy for these important markets," he noted.
During a question-and-answer session, Huang said that Nvidia's next investments in China could be in the medicine industry. Human biology is too complicated and diverse, he said in the interview. "No math equation will ever solve it. That's why we need a new form of technology… artificial intelligence.
"By infusing AI software into the workflow of scientists and doctors, I believe we will be able to do a better job: discovering the next drug or early detection of disease, even managing people who are living with disease. These are previously unsolvable problems that I believe are finally solvable, and that's one of the reasons why we're starting to invest in this area."
Huang was full of praise for China's big three tech companies, Alibaba Group Holding Ltd. [NYSE:BABA], Baidu Inc. [NASDAQ:BIDU] and Tencent Holding Ltd. [HKG:0700], which he notes are now "global" firms, not just Chinese enterprises. Nvidia has cooperated with all three, and they have adopted Nvidia's AI platform for internal development research and placed it on the cloud for Chinese startups to use. There are 300 startups in China using deep learning and Nvidia's platform very extensively, Huang added, saying that these companies now have access to infrastructure that only world-class companies could afford.
On investment in China, Huang talked of partnerships with leading universities and their groundbreaking work in artificial intelligence. AI startups in the country have access to technology, talent and money, he added.
There are so many startups. The spirit of entrepreneurs is very strong in China. And that is one of China's great national resources.