After the crisis, the boom. This is the rule of the market. And in 2024, once the chip crisis dictated by the Covid-19 pandemic has been overcome, the technological revolution of artificial intelligence erupts. Chips become the most precious raw material, capable of shaking markets and geopolitical balances. In this context of exponential evolution, Nvidia places itself as the undisputed protagonist. The company founded in 1993 by Jen-Hsun Huang to produce GPUs, graphics units for video gamers, has over the years become a leader in the production of semiconductors and chips for AI.
Let’s see the numbers. Nvidia ended 2023 with a 400% increase in profits, and in the second quarter of 2024 it continues to grow 265% compared to the previous year. The announcement of the new chip architecture for AI, Rubin, which will arrive in 2026, gives it a historic record in 2024: it is the most valuable listed company in the world. An (almost) unexpected overtaking of Apple and Microsoft, tech giants who have been competing for supremacy in the sector’s stock markets for years.
A flight that is anything but fancy, which finds firm roots in the evolution that the technological sector is experiencing post-Covid. A clear line between a past and a present that is yet to be written. The year of the pandemic will be remembered for the crisis of chips and in particular GPUs, essential IT components for which Nvidia enjoys fame as a global manufacturer for excellence and performance.
The cause is soon explained. Increase in demand which, in the pandemic crisis, was impossible to respond to. Remote working, distance learning and the growth of cryptocurrency mining are the three key factors: the demand for electronic devices is growing exponentially at the same time as production stops due to Covid. Factory closures in China and travel restrictions make it impossible to import the components needed to produce the devices. The result has been an increase in costs for GPUs and chips, with giants such as Apple and Sony forced to postpone the launch of new flagship products on the market, from the iPhone to the PlayStation 5.
Meanwhile, relations between the United States and China are becoming increasingly tense. The United States imposes severe restrictions on technology exports from China in an attempt to maintain a strategic advantage, with Taiwan, home of TSMC (Taiwan Semiconductor Manufacturing Company), the world’s largest chip maker, becoming a crucial node in the tech war that advance. The United States strengthens relations with Taiwan to ensure a stable supply of semiconductors, while China increases domestic production.
In 2022, Russia invades Ukraine. And it’s the start of a war that adds complexity and uncertainty to the semiconductor market. The sanctions imposed on Russia on energy supply chains and raw materials markets indirectly and inevitably also affect the semiconductor sector. Cooperation between Russia and China is growing, Putin’s strategy to circumvent Western sanctions, while the United States and its allies have strengthened measures to protect their technologies.
In a delicate geopolitical and technological advancement context, the new shock arrives in 2023 with the release of ChatGPT by OpenAI. The era of artificial intelligence opens, a completely painless revolution that shakes all sectors and markets. The United States and China once again find themselves fighting for technological supremacy, while Big Tech companies from all over the world compete with new AI solutions and new alliances. Microsoft vs Google. ChatGPT vs Gemini. Apple “allies” with OpenAI. The common denominator in the challenge is the need for increasingly powerful AI chips, to which Nvidia promptly responds with its DGX H100 in 2023, beating even its direct competitors Intel and AMD in terms of time and performance.
The difference between Nvidia and competitors in the design and creation of AI chips is the CUDA architecture (Compute Unified Device Architecture), a platform designed to guarantee the best performance of the large language models (LLMs) developed by OpenAI, such as ChatGPT , thanks to its ability to process huge amounts of data simultaneously. The combo between DGX H100 and CUDA proved to be a winner. On the one hand, it offers unprecedented computational performance. On the other, a memory optimized for AI and machine learning. For these reasons, Intel’s Xeon processors and AMD’s AI chips struggle to match it in performance.
For example, a well-designed CUDA architecture can handle training AI models much faster than traditional CPUs, reducing processing times from days to hours. A fundamental aspect in applications that require the processing of large amounts of data, such as image recognition, machine translation and autonomous driving. An aspect which, above all, justifies Nvidia’s rise among the tech giants.
As with every technological revolution and rapid rise in the market, there is no shortage of skeptics. There are those who consider Nvidia’s boom a potential market bubble. Doubts are legitimate and depend on two factors: the long-term sustainability of the demand for AI and the possibility that any technological or regulatory developments could slow its expansion. The market, while enthusiastic, remains cautiously optimistic, aware that even the most promising technological revolutions can encounter unexpected obstacles.
Huang’s giant plays its cards decisively to maintain its place at the top compared to other tech giants. With continuous investments in research and development and strategic partnerships, he proves himself to be not just a leader, but a pioneer.
As Amazon and Google expand their AI infrastructures and Elon Musk invests in advanced robotics and artificial intelligence through Tesla and SpaceX, Nvidia remains at the center of the global digital transformation, shaping the future of the technology industry and supporting the revolution with its performance devices AI that advances.