China No Longer Needs Taiwan for Chips

The semiconductor sector seems to be one of the last areas where China is still lagging far behind. The question today is whether China can close that gap or whether it will widen due to global export restrictions on chips. The answer isn’t as binary and linear as Washington will have us believe. The problem with this rather straightforward reasoning is that it ignores the regenerative nature of the chip sector.

If you listen to mainstream Western media, the battle lines are clear. The United States controls the choke points, ASML keeps its EUV lithography machines out of Chinese hands, and TSMC's factories in Taiwan remain the unreachable crown jewel. China has been encircled. And we all know the Chinese are masters of the game Go, where being encircled is the worst position you can be in. So the narrative goes that China will inevitably invade Taiwan to seize TSMC's factories, reinforced by Beijing’s more assertive language and military show of force around the island.

But if you take a closer look at what is actually happening in China today in terms of innovation and invention, a very different reality emerges. It is a reality I have extensively written about in my book China's Next Miracle. Today, China is not trying to catch up anymore in a race defined by the West. China’s race is run with hurdles, and China is clearing one after the other now.

When we analyze the latest innovations, funding, and infrastructure plans in China's semiconductor ecosystem, we see a clear, coordinated strategy unfolding. Not a sequential story where one chapter follows the next, but four fronts being pushed forward simultaneously across four parallel tracks. Together they explain exactly why an invasion of Taiwan for chips makes no sense.

Front 1: The Breakthrough Factory

The first front is about survival. Cut off from ASML’s Extreme Ultraviolet (EUV) lithography machines, China had to find ways to keep its tech companies running. So, for the past years, China executed its inverted innovation model, which I explain in detail in my book. Built pragmatically, driven by survival, finding "quick and dirty" solutions that work. Clearing hurdle after hurdle.

We see this in Huawei's recent patent for Self-Aligned Quadruple Patterning (SAQP). Instead of waiting for a domestic EUV machine, Huawei figured out how to use older, unrestricted DUV tools to etch patterns tight enough for 2-nanometer-class chips. It is expensive, it is inefficient, has lower yield, and it requires brute force. But it works. And remember, SMIC is already producing 5nm chips at scale today, powering Huawei's Kirin SoC processors, all without a single EUV machine from ASML.

At the same time, China is not betting on a single runner to crack the EUV problem. It is running a dozen parallel bets across the whole nation. We see SMEE filing patents for traditional lithography, Harbin Institute of Technology winning awards for its alternative laser-induced plasma methods, and startups like Hefei Lumiverse building desktop-sized EUV light sources. And behind closed doors, a "Manhattan Project" in Shenzhen is already assembling working EUV prototypes, it seems.

“Had the US not placed export restrictions on chips, Huawei would most likely have simply kept sourcing them from around the world, and the Chinese chip sector probably wouldn’t have evolved so rapidly into such a commercial threat to the US—which is exactly what’s happened.” — China's Next Miracle, p. 198

Front 2: The Leapfrog

The second front is where the story gets truly fascinating. While the West is obsessed with TSMC+ASML’s ability to shrink silicon transistors down to the last atomic layer, China is very quietly funding the next 50 years of computing. If you cannot easily win the nanometer race on silicon, why not change the material instead?

Readers of this newsletter have been following these breakthroughs one by one over the past three years, and what once looked like isolated Chinese lab curiosities is now forming a clear pattern. We are seeing the emergence of post-silicon technologies that bypass traditional lithography bottlenecks entirely.

Look at China’s Meteor-1 photonic chips, which use light instead of electricity to compute, covered in the newsletter in November 2025. Look at the world's first carbon-based AI chip, which uses ternary base-3 logic instead of the binary logic we have used for decades. Look at Fudan University's WUJI microprocessor, built on 2D MoS₂ materials, a 1 nm chip created without ASML machines, which I highlighted in December 2025.

"The battle for EUV machines could quickly become irrelevant if the underlying technology makes its own leap forward. If China never regains access to the latest chips and machines, it will simply continue to develop its own alternatives." — China's Next Miracle, p. 200

The most immediate and disruptive leapfrog is happening in advanced packaging. As Nvidia's Jensen Huang recently mentioned, export controls are failing because "chips for AI can be improved a lot with just architectural changes and packaging." He explains beautifully in this podcast how the AI industry has five layers (energy, chips, infrastructure, models, and applications) and that China is doing very well in all of the five layers, while the US is mostly fixated on one layer: the models.

Podcast April 15, 2025. China topics starts at 00:57 and talks about packaging at 1:06

Huawei’s Ascend 910C is the perfect proof of that concept. Unable to manufacture a single monolithic die that matches Nvidia’s H100, Huawei simply took two older 7nm chips and stitched them together using advanced 2.5D multi-chip packaging. By leveraging its 40% global market share in chip packaging, China has turned its current weakness in advanced lithography into a strength in systems engineering.

"The city of Wuxi, near Shanghai, is now home to 600 chip packaging companies and aims to become the Silicon Valley of chiplets. If China can’t make chips smaller, it will simply think bigger." — China's Next Miracle, p. 200

China is not alone in this new materials race. IMEC in Belgium, for example, is doing world-class research on 2D materials and carbon nanotubes. The choke points that strangled China in the silicon era, access to ASML machines, advanced EUV tools, and Western-controlled supply chains, simply do not exist yet in the post-silicon world. There are no export controls on MoS₂. There is no ASML for photonic chips. The playing field has not been built yet, and China is the one pouring the concrete and might even hold some new choke points themselves.

Front 3: The Chip Factory of the World

The third front is the one that will reshape the global industrial landscape. While all eyes are on the AI frontier and the most advanced chips, China is quietly building the factory of the world for everything else. According to SEMI and Allianz Research, China started construction on roughly 37 new chip fabs between 2023 and 2025. That is more than double the United States and Europe combined. Some of these projects failed, but the survivors are scaling production at an unprecedented rate.

By 2028, China is projected to control 42% of the world’s mature-node chip capacity. Jensen Huang even argued that China manufactures 60% of the world’s mainstream chips, maybe more. These are the 28nm and older chips that are not able to run large language models, but they do run our cars, medical devices, wind turbines, and most defense systems.

“Wars are no longer won solely by technological superiority, but increasingly by production capacity, scale, and speed—and this is equally true of the chip war. The battle for dominance in the semiconductor industry is no longer just about quality, but increasingly also about quantity. China’s chip industry is rapidly evolving into a kind of Temu platform for semiconductors: cheaper, delivered quickly, customized, and soon able to serve the majority of global chip demand.” — China's Next Miracle, p. 199

Front 4: The Taiwan Illusion

Western media frequently portrays a Chinese invasion of Taiwan as a chip war, a desperate grab for TSMC's advanced fabs. But this zero-sum game argument has a fatal flaw. The moment Chinese troops cross the strait, TSMC's chips stop shipping worldwide. The fabs do not come with an instruction manual. And ASML's machines? They come with a remote kill switch. Most TSMC engineers leave, and the supply chains will collapse. And the global economic fallout would dwarf anything we saw during the pandemic, including for China itself. An invasion for chips would destroy the very thing China is trying to acquire.

But when you understand the first three fronts of this story, you realize how that logic runs counter to China's strategy. Why would China risk an economic crisis, knowing that time will give them an advantage on three other fronts, just to seize factories now that rely entirely on Western equipment, software, and chemicals to operate?

"Why would China risk a catastrophic war that, would not only destabilize the region but also undermine China’s economic ambitions—not to mention reinforce the American narrative that China is aggressive and unreliable?"—China's Next Miracle p. 298

Through breakthrough factories (Front 1), technological leapfrogging (Front 2), and dominating mature nodes (Front 3), China is building an alternative semiconductor ecosystem that no longer depends on Taiwan. Combine that with Trump's visible disinterest in defending Taiwan's independence the way his predecessors did, and the picture shifts entirely. As Trump prepares to meet Xi Jinping in Beijing on May 14–15, China may not need a military invasion at all (Front 4). Time, technology, and diplomacy could do what tanks cannot: draw Taiwan quietly into China's sphere of influence. Imagine if Beijing makes a deal with Trump: TSMC in exchange for Taiwan. Crazy? Sure, but with Trump, no deal is inconceivable.

"China will never be able to replace all the world’s chip technologies and become fully independent, but with every new Chinese chip company supplying the world, our dependence on Chinese chips grows... Ultimately, it will be mutual dependence—and not absolute dominance—that will determine the new equilibrium." China’s Next Miracle, p. 199