NVIDIA Holds 80% AI Chip Market Share: Who's the Next AI Chip Supplier?

If there were a word to sum up NVIDIA CEO Jensen Huang’s response to growing competition in the AI chip segment, it would be “paranoid”, as he told CNBC in the first half of 2024. Despite a major stake in the market—high enough for the company to be referred to as a “moat”—the company still faces an impending boom in artificial intelligence (AI) chip creation. 

However, NVIDIA’s response is simply to accelerate its efforts, bringing new chip architecture to the industry every year—each more and more capable of training and deploying AI models—as opposed to its current biennial rate of technology introductions. So, is it possible for someone to take away some of that market share or will NVIDIA monopolize the AI chip market? 

NVIDIA’s Immense Growth and AI Chip Prowess

In H1 of 2024, NVIDIA’s growth rocketed with a 27% increase in share price, raising the company’s market cap to US$2.7 trillion. The main growth drivers for the business are its AI and data center businesses—where it excels in graphics processing unit (GPU) development for these rapidly-evolving industries. 

In terms of AI, though, we can see an increasing input as the company delved directly into emerging markets, such as electrification and autonomous mobility. Alongside NVIDIA’s groundbreaking developments, companies like Amazon’s Zoox, and General Motors’ Cruise (despite controversy) are making strides in the robotaxi sector. 

In terms of AI in data centers, the company tends to the global demand for greater computing power—a necessity for generative AI applications. Since 2021, NVIDIA has openly worked to 10x computing power, building Arm-based CPUs for faster servers with greater efficiency to enable complex AI programmes that leverage enormous datasets. 

So, NVIDIA is leading the way in the most crucial areas of innovation, and is growing exponentially as a result, but why would it be worried about competition?

Anticipating Subsidence of its First-Mover Advantage

If NVIDIA were to settle into its current seat at the top of the industry it would be a detrimental move to make, and the company’s CEO recognizes that. As it becomes one of the highest value businesses in the world, there is no time for complacency as its neighbors at the peak become the immediate risk factors—the likes of Alphabet, Microsoft, and Apple are already bolstering their positions through generative AI and the development of their own advanced silicon chips. 

Some of its biggest customers, including the aforementioned tech giants, are developing their own solutions in aid of cost reduction for themselves and their customers. 

Alphabet 

The owner of Google is one to watch as it competes in similar industries to NVIDIA. According to news released in May 2024, Google will 5x increase data center performance, which is a result of its tight-knit integration between its internal software and tensor processing units (TPUs). The search giant already commands the majority of the remaining 20% market share, but the recent announcement of its Trillium AI chip could help acquire a percentage of NVIDIA’s (estimated) 80%.

Microsoft

Towards the end of 2023, Microsoft Azure announced details of its very own silicon chips for AI—the Maia 100 and Cobalt 100. Largely beneficial for its cloud infrastructure, this cost-saving move will reduce the company’s dependence on NVIDIA chips. The alarming note for the chip leader is Microsoft’s backing of OpenAI, which is growing in value—currently reaching a US$80 billion valuation as a result of a major deal reported in February 2024.

Apple 

NVIDIA has been closing the gap with Apple as semiconductor chips and other AI-based software allow the company to rise to the top of the ranks. However, there is some speculation as to whether NVIDIA’s position will be short-lived as Apple intends to use its own silicon chips for its products. We know that Apple is very good at building hype around in its product launches, but also rapidly building upon today’s product offerings. It seems Apple, like other tech companies, is taking matters into its own hands by internalizing more of its product development. 

Amazon

Amazon’s Inferentia chip will power its products’ machine learning capabilities, including the much-loved Alexa voice assistant, which climbs towards an annual revenue of US$600 billion for the company. The act of shifting its processing power to internally-built components could eliminate a significant chunk of NVIDIA’s business as its current provider. 

What about NVIDIA’s main chip rivals Intel and AMD?

NVIDIA’s transition from gaming tech to data-center components gave the company a significant edge over its main rivals. Intel and AMD, although developing AI solutions of their own, represent a smaller, combined share of the market. 

In April 2024, AMD responded to mounting competition between the three legacy chip companies and brought out new AI computer processors. These parts will provide solutions to different functions capable of leveraging AI to enhance user experience, including the Ryzen PRO 8040 Series and the AMD Ryzen PRO 8000 Series desktop processors, which are built-in parts capable of powering AI-based functions. 

Intel will directly compete with NVIDIA in the data-center, edge and devices markets, having revealed its Gaudi 3 chip. The saving grace for Intel are other industry players like Dell, Hewlett Packard, and Supermicro who will be able to acquire their new components in the third quarter of 2024. 

Growth in the data-center industry will likely play a major role in positioning chip providers. The majority of new data-center demand will be centered around digital innovations incorporating AI-based features. While companies like Apple are looking for ways to upgrade the performance of their products to house AI functions, organizations like Intel and AMD will directly compete with NVIDIA in its core markets, backing cloud-based services and large datasets.

Cost-Saving and Control of AI-Powered Business

What were once major customers of NVIDIA will soon become competitors in a new realm of AI-influenced chip research and, in order to maintain its position, the leading firm must amplify its efforts to stay ahead of the curve in the data-center market, but the market share of device chips could go to the leading tech authorities. While big tech companies can learn from the developments at NVIDIA, it is unlikely they will settle to follow in its footsteps, but rather seek untouched avenues for their specific products. 

Device Cost Savings: In the process, companies will not only find new ways to deliver their own technologies internally, but minimize developmental costs by internalizing chip innovation. This is likely a response to a multitude of disruptions in the industry and  could have a profound impact as the likes of NVIDIA, Intel, and AMD lose a major share of trade as its core customers take matters into their own hands. 

AI for Data Centers: NVIDIA was a fast entrant into the data-center space, which will support the case for its AI components in the years to come. It will, however, lose a portion of its custom with the likes of Google as it adopts its own tech for its global data network. 

Trading regulation changes: Fears that China trade relations will increase the costs of acquiring new tech, especially semiconductors, it’s important for leading organizations like Apple, Alphabet, Microsoft, and Amazon to cut the ‘middlemen’ and build their own relationships with manufacturers.