Google's NotebookLM and the future of AI inference are attracting attention from Nvidia's CEO.

November 22, 2024

Nvidia's CEO, Jensen Huang, is showing strong interest in Google's NotebookLM. NotebookLM is attracting attention as a multifunctional AI tool for summarizing documents, generating questions, and generating voice conversations, and he describes himself as having "thoroughly mastered" this tool. Behind this is the evolution of AI technology and the importance of 'inference.' AI inference is the process of making practical predictions and decisions after training the model, and Huang positions this as the 'key to AI success'.
Nvidia considers the expansion of AI inference demand as a new growth opportunity and is accelerating the production of the latest GPU 'Blackwell'. The company has dominated the AI training market so far, but is anticipating the increasing demand in the inference field. Nvidia's movements leading the evolution of the AI market will be a significant guide in indicating the direction of technological innovation in the future industry sector.

NotebookLM offered by Google is an innovative platform that goes beyond just an AI tool. This tool, available for free, is notable for not only summarizing documents and generating questions, but also for supporting voice interaction generation. Nvidia's CEO, Jensen Huang, publicly stated that this tool efficiently utilizes the content of PDFs and papers, expanding the scope of AI application.
The essential value of this tool lies not only in its high information processing capability but also in the intuitive interface that allows users to manipulate a vast amount of information. NotebookLM is believed to contribute to streamlining information organization in companies and research institutions, and its demand is expected to increase even more in today's information-overloaded age. This could lead to a future where AI is deeply integrated into daily business operations and decision-making.

Jensen Huang pointed out that "inference" is an essential element in the evolution of AI. Inference is the process where AI models, based on training data, derive practical answers to real-world problems, allowing AI to evolve from a mere data processing device to a decision support system.
However, realizing inference requires vast computational resources, with high-performance GPU chips being indispensable. Nvidia anticipates an increase in demand in this field and is proceeding at full speed with the production of the latest GPU, 'Blackwell'. The company's efforts can be seen as a symbol of technological innovation to address the challenges brought about by the proliferation of AI inference.

In the AI market, Nvidia's presence is increasingly prominent. While the company has led the GPU market for training purposes, it is now gradually shifting towards the inference field. Jensen Huang mentioned that the increase in inference tasks will become a new source of revenue, especially expecting the sales of the 'Blackwell' GPU to support growth in the current quarter.
The situation where demand exceeds supply is evidence that Nvidia's strategy accurately captures the market's needs. Simultaneously, it suggests changes in industrial structures brought about by the expansion of the AI market. Nvidia's efforts are not only advancing AI technology but also offering new guidelines that other companies should follow.