Ruling the charts

AI is fuelling Nvidia's rise

The evolution of Nvidia from game-chip designer to the world’s most valuable company is not accidental. It is directly connected to its expertise in designing semiconductors that are good at intensive number-crunching. The timeline is straightforward. ChatGPT was publicly released in November 2022, sparking a frenzy of artificial intelligence (AI) development. In 2023, Nvidia more than doubled revenues to $60.9 billion as its chips became the hottest commodity in the world. Nvidia’s net profits expanded six times to $29.7 billion. The company’s market capitalisation crossed $1 trillion for the first time in May 2023 and it has tripled in the next 13 months. Nvidia is unusual in that it is a chip-maker that doesn’t make chips. It designs specialised semiconductors, which are manufactured to specification by contract fabricators. It is even open to having its chips manufactured by foundries owned by rivals like Intel.

Nvidia’s initial focus was the gaming market and it remains the dominant player with around an 80 per cent market share in gaming. Gaming created a demand for co-processors (also called graphics processing units or GPUs) to handle complex calculations for image-rendering, while leaving the central processing unit (CPU) free to handle less calculation-heavy tasks. AI requires similar but even more intensive processing by GPUs and Nvidia’s expertise has helped it become a leader in AI. Apart from expertise in chip design, the company has unusual characteristics in terms of corporate culture. It has a small workforce for a trillion-dollar corporation with fewer than 30,000 employees. Among other trillion-dollar concerns, Microsoft has more than 220,000 employees, while Apple has a headcount of over 160,000 and Google more than 180,000.

Its name derives from wordplay on “Next Version” and “Invidia”, which is the Latin word for envy. For many years, internal messaging from its President and Chief Executive Officer Jensen Huang always started with a warning: “Our company is thirty days from going out of business.” This was actually the case in 1997, when Nvidia was on the brink of folding up. The messaging embodies the sense of frenetic desperation that continues to drive Nvidia. It is now involved in a number of partnerships for research and deployment, including many partnerships in India, notably with Tata and Reliance Groups, as well as several of the Indian Institutes of Technology.

The US has blocked the export of certain Nvidia chips to China because of possible military applications. Some of its top-end GPUs, which cost $100,000-plus, are in such short supply that they are delivered under armed guards, and can be resold at significant premiums. The company has also been criticised by Linux developers (including Linus Torvalds) and by the gamer community for not open-sourcing drivers — the codes that run its chips. This makes it hard for developers and game designers to optimise Nvidia performance. However, Nvidia has bowed to community and peer pressure and started to open-source because its rivals Intel and AMD do so. Management theorists might cite one significant concern. Nvidia has risen to monopoly status, which is why it is valued so highly. It is such a critical component of the AI supply chain that a slight blip in its performance could derail the entire industry. Until such time as its rivals develop credible competencies, this would remain cause for worry.