The AI boom has made NVIDIA the dominant chip company in the world. But more and more rivals want a piece of the billion-dollar market.
• Despite NVIDIA’s dominance, the number of potential AI chip competitors is growing rapidly
• Large tech companies are developing their own AI chips
• Start-ups and Chinese companies are increasing the competitive pressure
The AI boom has made chip company NVIDIA the dominant player in the global semiconductor market. The company’s graphics processing units (GPUs) are currently the standard for training large AI models – and demand is growing rapidly. The group’s income also increases accordingly.
But the success has a downside: the enormous market power and the high cost of NVIDIA chips are driving customers and investors to look for alternatives. An increasingly dense network of rivals is emerging around AI hardware – from tech giants to start-ups to Chinese competitors.
Alphabet, Microsoft, Meta and Co.: Cloud giants are developing their own AI chips
According to Business Insider, the biggest challengers are NVIDIA’s most important customers. Large tech companies such as Alphabet, Amazon, Microsoft and Meta are investing heavily in their own AI hardware.
Google has been working on so-called Tensor Processing Units (TPUs) for around a decade. For a long time they were only used internally or in the Google cloud. But the group is now opening up its chips more to external customers. In February, Google entered into an agreement to lease its TPUs to Meta and is also working with cloud provider Fluidstack to lease them, reports Business Insider.
Amazon also relies on its own developments: The company offers Trainium chips for training AI models and Inferentia hardware for AI inference – i.e. executing models that have already been trained.
Microsoft and Meta are still at the beginning of their chip offensive, but are pushing it forward aggressively. Meta announced that it would develop four new generations of silicon in the next two years. The new AI model “Muse Spark” was only presented here on April 8, 2026. Microsoft, on the other hand, recently presented the Maia 200 inference chip.
AI chip boom: start-ups are getting involved
In addition to the tech giants, specialized start-ups are increasingly getting involved, as Business Insider emphasizes. Investors are pumping billions into young companies focused on inference chips. This area is considered particularly cost-sensitive because AI models have to be operated here permanently.
Business Insider cites Cerebras Systems as a prominent example. Founded in 2015, the company builds huge “wafer-scale” chips that can be used for both training and inference. Cerebras also closed a $10 billion deal with OpenAI in January.
SambaNova Systems and Tenstorrent are also positioning themselves as alternatives to classic GPUs, it goes on to say. SambaNova recently raised $350 million after takeover talks with Intel failed.
Even NVIDIA is watching this trend closely: As Business Insider reports, the company paid around $20 billion to license technology and recruit top talent from AI chip start-up Groq.
Huawai and Co.: Geopolitics and China as a strategic factor
Further competitive pressure arises from geopolitical developments. The USA has tightened export restrictions on powerful AI chips to China. At the same time, Chinese companies are working intensively on their own alternatives.
According to Business Insider, the focus is on the telecommunications giant Huawei, which develops chips, servers, network hardware and cloud services from a single source. According to some experts, Huawei is therefore considered a potential Chinese counterpart to NVIDIA. But companies like Cambricon Technologies and the Internet companies Alibaba Group and Baidu are also developing their own AI chips for their cloud businesses.
AMD, Intel, Broadcom and Co.: Established chip giants are also attacking
In addition to new competitors, established semiconductor manufacturers are also trying to benefit from the AI boom. These include in particular AMD, Intel and Broadcom, as the Business Insider report shows. AMD positions its GPUs as a direct alternative to NVIDIA hardware and has already won contracts with major cloud customers such as Meta. Intel, on the other hand, has a strong position among enterprise customers. Broadcom is focusing more on networking and custom chips – and could benefit even if NVIDIA remains the leader in the GPU segment.
Despite NVIDIA’s continued enormous lead, competition in the AI chip market is becoming more intense. Cloud giants are developing their own hardware, start-ups are relying on specialized solutions and geopolitical tensions are promoting regional alternatives. NVIDIA is still clearly leading the field – but the number of challengers is growing rapidly.
Bettina Schneider, editorial team at finanzen.net
This text is for informational purposes only and does not constitute an investment recommendation. finanzen.net GmbH excludes any claims for recourse.
By the way: Alibaba and other US stocks can even be traded on finanzen.net ZERO until 11 p.m. (without order fees, plus spreads). Open a depot now for free and secure a new customer bonus!
Selected leverage products on Alibaba
With knock-outs, speculative investors can participate disproportionately in price movements. Simply choose the lever you want and we will show you suitable open-end products on Alibaba
The leverage must be between 2 and 20
Advertising
Image sources: Konstantin Savusia / Shutterstock.com, Below the Sky / Shutterstock.com
