The current AI wave has triggered a frenzy in the tech sector. Companies recognise that LLMs may make natural language the chief interface for computer interaction. Hence, it’s no surprise that tech companies are in a hurry to develop their own Large Language Models. For this, they need hardware. In the case of AI, a particular kind of hardware is vital: an AI chip.
Meticulously designed, this chip is capable of rapidly processing data-heavy tasks with greater efficiency than conventional computer chips. The snag, however, is that it has now become the primary bottleneck in AI development.
Lisa Su, AMD’s chief executive, forecasts a surge in the AI chip market from $30 billion this year to an impressive $150 billion in 2027.

One company in particular — NVIDIA — has a distinct advantage in the market. In this Chartpack, we cover:
How NVIDIA built a moat in the AI chip market,
What the competitive landscape tells us about the industry’s near future,
And how open-source AI development could challenge NVIDIA’s dominance.
Let’s go!
One chip (company) to rule them all
In recent years, NVIDIA has become the undisputed leader in AI chips. In academic papers, NVIDIA chips are overwhelmingly the most referenced, accounting for a whopping 94% of AI papers that specified the use of a particular chip in 2022.

Keep reading with a 7-day free trial
Subscribe to Exponential View to keep reading this post and get 7 days of free access to the full post archives.