Huawei is building an AI chip set to launch in early 2027, and it could crack the one market Nvidia thought it owned permanently. China’s biggest tech company isn’t slowing down under US sanctions. It’s accelerating. If Huawei pulls this off, the $500 billion AI chip market will never look the same.
Why This Is Happening Now
The US government has spent three years trying to cut China off from advanced semiconductors. Export controls banned Nvidia from selling its H100 and A100 chips to Chinese buyers, according to the US Commerce Department. Beijing responded the way it always does: it funded its own.
Huawei’s Ascend chip series has been quietly gaining ground inside China’s largest AI firms. According to Reuters, Baidu, ByteDance, and dozens of AI labs backed by the state have already shifted some workloads from Nvidia hardware to Huawei’s Ascend 910B. The shift is small today. It won’t stay small.
Now Huawei is preparing its next generation chip for a Q1 2027 launch, according to the South China Morning Post. The new chip is expected to target the same training and inference workloads that made Nvidia’s H100 a chip every AI lab in the world had to have.
Nvidia’s data center revenue hit $47.5 billion in fiscal year 2024, according to Nvidia’s annual report. That number tells you everything. This is the market Huawei is walking into.
The Contrarian Take Nobody Wants to Say Out Loud
Most analysts are asking whether Huawei’s chip can match Nvidia’s performance. That’s the wrong question.
The right question is: does it need to?
China’s AI market isn’t going to keep waiting for Nvidia chips it can’t legally buy. Chinese firms are adapting their models and software stacks to run efficiently on Huawei hardware. According to Bloomberg Intelligence, China’s AI chip demand is projected to reach $27 billion annually by 2028. Huawei doesn’t need to win globally. It just needs to win at home first.
This is where the rich versus poor mindset splits people into two camps.
The average investor sees “Huawei chip” and thinks “inferior knockoff.” They dismiss it. They stay long Nvidia because Nvidia is “the winner.” They never ask what happens to Nvidia’s addressable market if China builds a wall around its own AI spending.
The sharp operator sees a second supply chain forming. They understand that when governments force decoupling, you don’t get one winner. You get two markets. Two chipmakers. Two separate tech stacks. And the companies that figure out how to sell into both, or hedge across both, will print money while everyone else picks sides.
Nvidia is not going away. But its growth story assumed it would dominate China’s AI buildout. That assumption is now cracking. According to analysts at Bernstein Research, China represented roughly 20 to 25 percent of Nvidia’s data center sales before export controls tightened in 2023. That revenue has to go somewhere. Huawei is building the place it goes.
If you’re running a tech business right now and you have exposure to hardware costs, AI infrastructure, or procurement from Asian suppliers, this shift matters for your cash flow and your vendor risk. Smart operators are already mapping their supply chain dependencies. Tools like Wallester’s business card platform make it easier to track spending by vendor category, so you can see exactly how much of your tech budget flows through single points of failure before they become crises.
What This Means for You
Here is what I would do if I were running a company with any AI infrastructure exposure.
First, stop assuming the chip supply chain is stable. It is not. US policy can shift, Chinese manufacturing can scale faster than anyone expects, and Nvidia’s pricing power depends entirely on having no real competitor. A credible Huawei chip changes that math. Not overnight. But it changes it.
Second, watch where Chinese AI firms are investing. ByteDance, Alibaba, and Tencent are not hobby projects. They’re among the most sophisticated AI operators in the world. If they are integrating Huawei chips into their production stacks, that tells you the hardware is good enough to ship real products on. That’s the real benchmark, not scores posted on tech news sites.
Third, think about talent. The chip race between the US and China is pulling engineers, researchers, and chip architects in both directions. Hiring in AI hardware or semiconductor design is about to get more expensive. If you’re building a team in this space, get your compensation structure right before the bidding war gets worse. Gusto’s payroll platform is worth using here because it gives you clean reporting on total compensation costs as you scale technical headcount, which matters when investors or partners ask how you’re managing burn.
Fourth, pay attention to the software layer. The company that makes it easiest for developers to write code that runs on both Nvidia’s CUDA framework and Huawei’s CANN framework will capture enormous value. That abstraction layer is worth billions. Someone is going to build it. The question is who gets there first.
The Bottom Line
Huawei is not building a chip to impress American tech journalists. It’s building a chip to take back $27 billion in annual AI spending that US sanctions handed to domestic Chinese manufacturers. The Q1 2027 launch is not a moonshot. It’s a business move backed by the full weight of the Chinese state. Nvidia now has a real competitor on its flank. The only question is whether you see that as a threat or an opportunity before the market prices it in.
Frequently Asked Questions
What AI chip is Huawei planning to launch in 2027?
Huawei is preparing a next generation AI chip targeted at training and inference workloads, set for a Q1 2027 release according to the South China Morning Post. The chip is expected to build on the Ascend 910 series, which Chinese AI firms have already begun adopting at scale for production workloads.
Can Huawei’s AI chip really compete with Nvidia?
For global markets, probably not immediately. But for China’s domestic AI market, the standard is different. Chinese firms need chips they can legally buy and integrate without worrying about US export enforcement. Huawei’s Ascend hardware is already running production workloads at companies like Baidu and ByteDance, according to Reuters. Good enough to ship is good enough to take market share.
How do US export controls affect the Huawei AI chip launch?
US export controls banned Nvidia from selling its most advanced chips to Chinese buyers, according to the US Commerce Department. Those restrictions have been tightened multiple times since 2022. The result is that Huawei’s chip program now has guaranteed domestic demand before it even ships, because Chinese AI labs have no other viable option at scale.
What does this mean for Nvidia’s revenue?
According to Bernstein Research, China accounted for 20 to 25 percent of Nvidia’s data center revenue at its peak before export restrictions tightened. A credible Huawei competitor formalizes that loss and caps Nvidia’s China upside for the foreseeable future. Nvidia’s global dominance stays intact for now, but the total addressable market it can actually reach just got smaller.
Should investors pay attention to the broader Chinese semiconductor sector?
Yes, but with eyes open. The domestic chip investment wave in China lifts firms across the supply chain, from foundries like SMIC to chip design companies like Cambricon. The risk is real: US and China policy swings can move these valuations hard in either direction, often overnight, so position size matters more than conviction here.
