About a week ago, a viral post declared that on July 19, China had “killed the silicon wafer.”
The claim was explosive: a breakthrough in a new semiconductor material called indium selenide (InSe) had supposedly rendered the entire Western chip ecosystem — from Intel’s FABs to TSMC’s foundries and America’s sanctions — obsolete overnight.
China, the post argued, had not just won the chip war; it had “exited the battlefield” by mastering a new law of atomic physics.
Like many things on the internet, this narrative was a dramatic oversimplification. However, it was pointing to a significant real-life event. An innovative method for the mass production of InSe wafers of high quality was described in an article that was published on July 18 by researchers from Peking University and Renmin University in China. While this achievement won’t kill silicon tomorrow, it represents a genuine strategic leap. It suggests that China is actively working to develop the next technological paradigm while the West has been concentrating on preventing the implementation of the current one. A more nuanced but equally profound narrative about the future of technology, geopolitics, and the very materials that will power our world emerges when the accuracy of the viral claims is compared to the scientific reality. Let’s examine the implications of this breakthrough for China, the West, and the semiconductor industry’s future. Then, we’ll close with my Product of the Week: the HyperX QuadCast 2 S microphone.
The Inside Story of China’s First InSe Wafer One thing is correct in the core of the social media post: stoichiometry, a problem of atomic-level precision, has been the primary obstacle in the production of InSe. InSe is a two-dimensional (2D) material, meaning it can form stable layers just a few atoms thick. An ideal one-to-one atomic ratio of indium and selenium is required for it to function as a high-performance semiconductor. Defects that ruin its electronic properties result from any deviation. In contrast to silicon, which can be polished and doped into submission and is a robust, forgiving element, InSe is unforgiving. The achievement of the Chinese scientists was cracking this very problem.
Their innovative method involves heating amorphous InSe film and solid indium in a sealed environment. In a self-correcting process, the vaporized indium produces a liquid interface that enables the formation of high-quality, atomically perfect InSe crystals. Crucially, they scaled this from microscopic lab flakes to 5-centimeter wafers and built functional transistor arrays, proving the material is “fabrication-grade.” This is a vital step in moving a material from the lab to the factory, a bottleneck that has stalled many promising post-silicon candidates.
Promise of a ‘Golden’ Semiconductor
The hype surrounding InSe is legitimate. As silicon-based chips shrink toward their physical limits, the industry is desperately searching for alternatives to continue the progress defined by Moore’s Law — the trend that Intel co-founder Gordon Moore identified in 1965, predicting transistor counts would double roughly every two years with minimal cost increases.
For a number of reasons, InSe, also known as a “golden” semiconductor, has long been a leading contender. First, compared to silicon, its electron mobility—the speed at which electrons move through it—is far superior. Some studies have shown that it can exceed 1,000 cm2/Vs. The material’s high electron mobility translates directly to faster switching speeds and more powerful processors.
One report suggests that transistors made with this material could triple the intrinsic switching speed of current 3nm silicon technology while improving energy efficiency by an order of magnitude.
Second, InSe is a true semiconductor with a tunable bandgap, making it suitable for digital logic, in contrast to the “wonder material” graphene, which has no natural bandgap and cannot be easily “switched off.” Finally, its atomically thin nature allows for superior gate control, mitigating the “short-channel effects” that plague modern silicon transistors and cause power leakage.
Fact Check: Why Silicon Is Still Around Here, however, is where the viral post veers from science into science fiction. It is fundamentally incorrect to assert that this breakthrough renders ASML, TSMC, and the entire Western supply chain irrelevant. The first, though crucial, step is making a flawless wafer. The magic of a modern chip lies in patterning trillions of transistors onto that wafer with nanometer precision.
This process depends on a mind-bogglingly complex and expensive ecosystem that China has yet to master domestically. It includes extreme ultraviolet (EUV) lithography machines — a particularly challenging gap — as well as advanced etching and deposition tools from companies such as Applied Materials and Lam Research. The new InSe growth method does nothing to reduce this reliance on a sanctioned, highly intricate infrastructure.
Furthermore, the global semiconductor industry is a multi-trillion-dollar behemoth built on a silicon-based infrastructure that has been optimized over the past 50 years. Although the 5cm InSe wafer is a stunning proof of concept, it is vastly superior to the 300mm (12-inch) wafers that are the norm in contemporary fabs. Transitioning the industry to new material will take decades and trillions of dollars, with significant challenges in yield, cost, and reliability. Silicon will remain the workhorse of the digital world for the foreseeable future.
A strategic leap, not a blowout. The true significance of this breakthrough is not that it renders silicon obsolete, but that it provides China with a powerful, sanction-proof path to developing next-generation technology for key strategic sectors where performance is paramount, and cost is secondary:
Military and Aerospace: This is the most immediate and critical area of impact. InSe-based chips may provide a significant performance advantage for satellite communications, advanced radar, and electronic warfare systems. By developing a sovereign capability in a “beyond-silicon” material, China can build specialized, high-end military hardware that isn’t dependent on U.S. technology.
AI, Cloud, and High-Performance Computing: The most significant commercial threat is to the dominance of companies like Nvidia. An important metric in the energy-hungry data centers that power the cloud and AI revolution is performance per watt, and an InSe-native AI accelerator may provide superior performance. Chinese companies may be able to build powerful, efficient, and entirely domestic AI supercomputers thanks to an InSe-native AI accelerator, which is a major national priority. Medical and consumer electronics: In the long run, InSe’s distinctive properties open up new product categories. Its superior flexibility makes it suitable for foldable displays and wearable electronics, and its sensitivity and low power consumption make it ideal for advanced medical sensors and Internet of Things (IoT) devices. InSe’s flexibility, sensitivity, and low power use create a market where China could leapfrog existing technologies and establish a new frontier of innovation.
Impact on Existing Chip Companies: While the incumbents aren’t out of date, they should take note. TSMC and Intel must accelerate R&D in 2D materials to avoid being outflanked. Equipment makers like ASML will still find a market, as future InSe fabs will need their lithography tools. The real losers are the sanctions, designed to trap China in the silicon paradigm, which it now has a credible path to sidestep.
Wrapping Up
The dramatic claim that China “killed the silicon wafer” is a wild exaggeration. However, a crucial truth lies beneath the exaggeration: a significant scientific breakthrough has taken place, and its geopolitical repercussions are significant. China has not ended the chip war, but it has successfully opened a new front — one fought not with geopolitics and supply chains, but with fundamental materials science. This achievement is a clear signal that a strategy based solely on containing an adversary’s access to existing technology is doomed to fail.
Whoever can master the atomic details of the materials that will replace silicon will not only determine the future of computing, but also who can etch the tiniest lines on it.
