Jensen Huang, CEO of NVIDIA, recently participated in an interview with the firm Altimeter Capitalwhere he talks openly about his company’s dominant position in the AI market. One of the statements that drew the most attention within the round of talks was that there are currently $1 trillion in data centers based on older CPUsand that all of them they will have to switch to GPU in the next 4 to 5 years.
Machine learning in everyday applications
It is difficult to estimate what the total investment amount would be during this time if companies choose to modernize their data centers with GPUs, but Huang is confident that this will be done, also using its products. As expected, the director of the technology company with the highest AI revenue in the world anticipates that these artificial intelligence processes will be present in “almost everything” and that it is inevitable.
“Each of the applications – Word, Excel, PowerPoint, Photoshop, Premiere, AutoCAD – tell me which application is your favorite… I assure you it will use a lot of machine learning in the future,” said the executive. This argument, he assures, is a fundamental reason for the modernization of 1 trillion dollars in data centers with his cards.
“Moore’s Law has largely ended.”
Jensen Huang on why Nvidia will replace ONE HUNDRED PERCENT of the $1 trillion CPU-based data center TAM with GPU.
“We’ve reinvented computing, we’re not going back.
“Imagine you have $50B of CapEx to spend:
“Option A, Option B; build CapEx… https://t.co/NG9KgX5XCZpic.twitter.com/xbXqG5uKYC
— Compound248 💰 (@compound248) October 13, 2024
Critical Considerations
Huang affirms that the company’s current optimism bears no similarity to the phenomenon that Cisco aroused at the height of the dotcom bubble and defends that they are “reinventing computing”, since the future will be “highly dominated by machine learning.” Meanwhile, the flow of its coffers does not stop with the production of its latest Blackwell GPUs sold out for the next 12 monthsas demand for those chips continues to be “exceptionally high.”
However, some questions are being raised about the future of generative AI and its ability to truly justify the investment. Critical sectors point out that the transition from CPU to GPU may not be viable for all companies, and that not all generative AI use cases will prove to be profitable. Despite this, Huang believes that every new investment dollar should be spent in this area.
Impact of Generative AI on business strategies
The director of NVIDIA cites the example of a company with 50 billion dollars in CapEx (capital for investments in infrastructure and equipment). He makes a distinction between “past” and “future” investment, arguing that it is entirely channeled into generative AI. “And now his company is better,” he says.
Jensen Huang’s vision for modernizing data centers and integrating AI into all applications reflects an ambitious and exciting future. While there are uncertainties about the viability of generative AI and the hardware transition, there is no doubt that your company is positioned as a leader in the sector, driving the economy into a new era of computing. The key will be how companies will be able to adapt their strategies and resources to take advantage of these innovative tools.