OpenAI raises $7 trillion to compete with Nvidia? Huang Renxun: This money can buy all GPUs
The Paper reporter Hu Hanyan
As stock prices hit new highs, Jensen Huang, founder and CEO of AI (artificial intelligence) chip giant NVIDIA, once again expressed his views on the development of chips and AI.
On February 12, local time, NVIDIA founder and CEO Jensen Huang stated at the World Government Summit (WGS) in Dubai that countries need to build independent AI infrastructure. Technological advances in the next few years will significantly reduce the cost of developing AI.
Huang Renxun believes that “the end of general computing and the beginning of accelerated computing” are ushering in, and NVIDIA is rapidly improving the efficiency of AI computing to achieve the “democratization” of using AI; each country needs to have its own sovereign AI to develop economic potential and preservation of national culture.
Shares of Nvidia (Nasdaq: NVDA) have soared nearly 50% this year. On February 12, NVIDIA’s stock price closed at US$722.48 per share, up 0.16%, and its total market value reached US$1.78 trillion, approaching Amazon (US$1.79 trillion) and Google’s parent company Alphabet (US$1.83 trillion). More than Amazon. The market expects that based on such a rise, Nvidia may soon become the world‘s third-largest technology giant with a market value after Microsoft ($3.09 trillion) and Apple ($2.89 trillion).
Recently, foreign media reported that Sam Altman, the founder and CEO of the emerging AI giant OpenAI, is raising a total of up to 7 trillion U.S. dollars in funds from the Middle East to support a company’s semiconductor plan and is working with Nvidia compete.
In response to this, Huang Renxun responded somewhat sarcastically at the meeting, saying that he believed that the chip industry will drive down the cost of AI: “(Seven trillion US dollars) can obviously buy all the GPUs… If you think computers can’t develop faster, One might conclude that we would need 14 planets, three galaxies and four suns to fuel all this. But computer architecture is constantly improving.”
However, Huang Renxun also emphasized that computing and AI technology are still accelerating, and related expenditures are growing: “The total value of global data centers is currently about one trillion U.S. dollars. In the next four to five years, this number will grow to two trillion U.S. dollars.” Trillions of dollars, these data centers will become the source of power for global software operations… In the past decade, one of our greatest contributions has been to advance computing and AI a million times.”
When it comes to AI safety issues, Huang Renxun believes that some concerns are exaggerated. AI can be regulated like industries such as automobiles and aviation, and praised open source models such as Llama 2 and Falcon for their contribution to AI safety: “Some people deliberately scare Everybody, to make people fearful of this new technology, to demystify it, to make other people stop taking action and do it all on their own. I think that’s wrong.”
Finally, Huang Renxun also shared his views on education, believing that AI has greatly reduced the technical barriers to programming, so he recommended that today’s young people learn “the most complex science to understand”, that is, biology: “It not only covers The content is broad and complex and difficult to understand, and the key is to have a huge impact… Every year, our software, chips and infrastructure develop better than the year before, but in the life sciences Little progress has been made.”