SMIC Begins Testing China’s First Homegrown Immersion Lithography Machine

China’s largest chipmaker, SMIC, has begun testing the first domestically assembled immersion lithography machine – a development that could mark the start of a new era for the nation’s semiconductor industry. While global leaders like TSMC and Samsung have been relying on advanced extreme ultraviolet (EUV) systems from ASML to push process nodes down to 2nm, China has long been constrained by U.S.

sanctions that ban the sale of these machines. For years, speculation swirled around whether Chinese companies would manage to build their own next-generation lithography equipment, and now a real-world attempt is beginning to take shape.

The machine in question is a deep ultraviolet (DUV) immersion lithography tool, developed by Shanghai Yuliangsheng Technology Co., a company with ties to Huawei. According to reports, SMIC is already conducting tests with this system, which has been built with the goal of minimizing reliance on foreign suppliers. While some parts are still imported, the company hopes to eventually source all components from within China – a move that would sidestep sanctions and allow Chinese foundries greater independence in the supply chain.

To understand the significance, it’s worth stepping back. Lithography machines are essentially the backbone of chip production. They work by projecting patterns of light onto silicon wafers, carving out circuitry at nanometer scale. In conventional lithography, a thin layer of air sits between the lens and the wafer, which slightly reduces resolution. Immersion lithography, however, swaps this gap for ultra-pure water, raising the numerical aperture and enabling finer, more precise lines to be etched. This technological leap is crucial in producing chips with billions of transistors, the engines that power everything from smartphones to artificial intelligence systems.

The Yuliangsheng DUV currently produces chips at the 28nm node in single exposures. While 28nm is far from cutting-edge – ASML first launched comparable immersion systems like the Twinscan NXT:1950i back in 2008 – the Chinese machine represents a vital proof of concept. With multiple patterning, this DUV system could theoretically allow SMIC to push down to 7nm or even 5nm production. Yet, this path is not without challenges. Multiple patterning increases costs, introduces risks such as overlay errors (misalignment of layers), and can result in pitch walking, where spacing between features becomes inconsistent. These problems directly reduce yield and raise the per-unit cost of chips.

Despite these hurdles, the long-term ambition for China is clear: developing a homegrown EUV system, ideally a High NA EUV, which would give its foundries the ability to mass-produce chips at 2nm and beyond. Such a breakthrough would allow SMIC to compete more directly with TSMC and Samsung, which are currently in a league of their own. At present, though, China is still at the stage of proving that it can build reliable immersion DUVs, a stepping stone on a long and complex journey.

The urgency of this development is underscored by Huawei’s experience. In 2023, the company stunned the tech world by unveiling the Kirin 9000S chipset inside the Mate 60 Pro smartphone. Manufactured at SMIC’s 7nm process node, this chip not only marked Huawei’s first high-performance in-house silicon in years but also restored native 5G capability to its flagship devices – something missing from the Mate 50, P50, and P60 lines, which were forced to use altered Qualcomm Snapdragon chips stripped of 5G functionality due to U.S. restrictions. While the 9000S was a technical achievement, sustaining and improving such production without domestic lithography tools will be extremely difficult.

For now, industry watchers estimate that SMIC could begin using the Yuliangsheng DUV machine in actual production by around 2027, starting with 28nm chips. This would still be a generation behind leading-edge manufacturers but represents a vital milestone. The real challenge will be scaling up the technology, refining manufacturing yields, and eventually moving toward more advanced nodes without relying on imported systems.

In essence, SMIC’s test signals both a limitation and a breakthrough: limitation, because China remains years behind the very top foundries; breakthrough, because it demonstrates that domestic innovation can provide a pathway around the technological blockade. The coming years will reveal whether these early steps can evolve into a fully independent lithography ecosystem capable of delivering true next-generation chips. If successful, China’s foundries could eventually narrow the gap with global leaders and reset the balance of power in the semiconductor industry.