It is estimated that Nvidia will lose around $16 billion in revenue this year; and the stock price continues to feel the pressure. This isn’t just about Nvidia, but as a leading designer of AI chips, emerging as a central figure in this geopolitical game of chess. There is of the course a looming shadow. DeepSeek’s low-cost AI model, released much to everyone’s surprise in January, upturned conventional AI model development methodology and cost structures. In following days, around $1 trillion was wiped off the United States (US) tech stocks.

Also Read: AI could pass any human exam in 5 years, clear medical tests: Nvidia CEO

Not only was this a concern, it heightened the US perception about China’s as-yet-unclear AI capabilities. Their idea, is to not strengthen it any further.

TSMC, or Taiwan Semiconductor Manufacturing Company, is believed to be hiking prices of the chips being sold in the US by as much as 30%. That cost will likely be passed to customers — your next tech purchases may become more expensive, gradually. You may wonder why prices of chips being manufactured in the US going up, since this wouldn’t come within the ambit of the tariffs. Many reasons for this — market demand and supply, costs at the Arizona fab is much higher compared with Taiwan, while certain pieces of the fabrication puzzle that still need to be imported.

To give you an idea of the interconnected nature of global markets and supply chains, we must weigh the importance of China for Nvidia, and of the US for TSMC. An illustration, to explain the bigger, complex picture.

Chipmakers, sales and AI development

China has historically accounted for a substantial portion of Nvidia’s revenue, with estimates suggesting it was a $20 billion market for the company despite already reduced exposure (that would be 15% of revenue in 2023). The ban on H20 chip sales alone is expected to cost Nvidia $5.5 billion in the near term. This financial strain comes at a time when Nvidia is investing heavily in new architectures like Blackwell, which could be further disrupted if export controls tighten.

Nvidia’s chips, such as A100, H100, A800, H800, and H20, are critical for powering AI applications, including training large language models, on massive amounts of data.

The US is TSMC’s largest market, accounting for a significant portion of its revenue due to the high demand for advanced semiconductors in AI, high-performance computing, smartphones, and other technologies. Apple, Nvidia, AMD, Qualcomm, MediaTek and Broadcom, some of their biggest customers. TSMC’s global revenue in 2023 was around $68 billion, of which the U.S. market generated around $44.2 billion.

TSMC’s Arizona fabs, producing 5nm and 4nm chips, cater to the US clients to mitigate supply chain risks.

In the long-term, Nvidia would bet on global demand for AI chips remain on an upward trajectory, and their dominance with chips meaning new market opportunities will open.

Also Read: Nvidia scientist praises intern who quit chipmaker to join DeepSeek: ‘Much impressed by his decision’

For Chinese companies, the tariffs, alongside Nvidia chip ban and a general restriction to accessing American tech, is a double-edged sword. It severely limits access to cutting-edge AI chips, which are essential for training large-scale AI models. Chinese tech giants including Alibaba and Huawei, as well as AI startups such as DeepSeek, are believed to have relied on Nvidia’s chips to find their competitive edge. This may be a big opportunity for Huawei, to tap the space that’s been vacated.

These constraints could slow China’s AI progress. On the other hand, this ban will almost certainly serve to accelerate China’s push for semiconductor self-sufficiency. There has been success. Huawei developing the Ascend 910 processor as a rival to Nvidia’s chips, is an example. There will be a demand to prioritise domestic AI chips over imported chips.

This will hurt everyone — Intel, Qualcomm, and AMD, come to mind.

Bifurcation, division and complication

The Nvidia chip ban is not an isolated event but a microcosm of the US-China tariff war’s broader dynamics. The tariff escalation, with US duties on Chinese goods jumping as high as 145% and now 245%, and with China retaliating with 84% and then 125% tariffs, has disrupted global supply chains and heightened economic uncertainty. Semiconductor companies face increased risks as China pushes for domestic production to achieve a target of 70% semiconductor self-sufficiency by 2025.

This could inevitably lead us to a bifurcated global chip market, and equally fragmented and bifurcated hardware as well as software ecosystems emerging: one led by the US and its allies, and another with China at the centre.

Outlandish? Not exactly. We’ve seen it before with a Google-led Android ecosystem, and an increasingly capable Huawei-led alternative that’s Google Services free. Huawei Mate XT Fold tri-fold smartphone is an achievement on the hardware and software collective, that no other phone maker has been able to replicate.

Look at how the Chinese electric vehicle (EV) makers such as Build Your Dreams (BYD) are developing, and even Tesla’s pursuits pale in comparison. Others, have a lot of catching up to do. That’s tech, and future mobility, covered.

All of this serves to also complicate supply chains for companies like Apple, Samsung, Intel, and many others, which rely on both US and Chinese markets. Not many may have realised, but China’s retaliatory measure to control exports of rare earth elements, will be troublesome for the production of pretty much any chips or electronics outside of China. There really cannot be a winner in all this.

Vishal Mathur is the technology editor for HT. Tech Tonic is a weekly column that looks at the impact of personal technology on the way we live, and vice-versa. The views expressed are personal.