introduction
For the chip manufacturing industry, lithography technology is undoubtedly a crucial core process. As Moore's Law continues to drive, the chip manufacturing process continues to shrink, and the requirements for lithography machines are becoming more and more demanding. From ordinary dry DUV lithography machines, to immersion DUVs, EUVs, and the increasingly expensive High-NA EUVs, technological advancements in the industry have repeatedly promoted the upgrading of lithography equipment.
However, such a technical path also imposes a heavy cost burden on chipmakers. TSMC, for example, already has hundreds of first-generation EUV lithography machines, with a total cost of tens of billions of dollars. In the future, the production of chips of 2nm and below will require more expensive High-NA EUV, and the cost will further rise.
Against this backdrop, TSMC recently announced a decision that shocked the industry: it decided not to use the expensive High-NA EUV, but to continue to use existing old EUV lithography machines to make 2nm and more advanced chips. This move is undoubtedly "flipping the table" and has had a wide impact on the entire industrial pattern with lithography machines as the core.
So, why did TSMC make such a choice? What kind of innovative technology is it able to achieve the 2nm process on the old lithography machine? What impact will this approach have on the chip industry in the future? Let's dive into this fascinating technology topic.
The continuous upgrading of lithography technology and the cost burden
As the core process of chip manufacturing, the development process of lithography technology can be said to be synchronized with Moore's law. From the original dry DUV lithography machine, to the immersion DUV, EUV and even High-NA EUV, the introduction of each new generation of technology means that the chip process has been scaled down again and the performance has been greatly improved.
Take TSMC as an example, the company has already started to use EUV lithography machines in the 7nm process. With the continuous evolution of the process, TSMC will undoubtedly need more advanced lithography technology support in the 5nm, 3nm and even future 2nm processes.
However, behind this technological advancement, it is also accompanied by huge capital investment. The cost of an ordinary DUV lithography machine is about 10 million US dollars. The immersion DUV lithography machine costs 4-50 million US dollars, the EUV lithography machine is hundreds of millions of dollars, and the High-NA EUV is as high as 300 million US dollars.
For a giant like TSMC, these costs are undoubtedly affordable. But the problem is that the conversion of each generation of process means a comprehensive replacement of the corresponding lithography machine. Taking EUV lithography machines as an example, TSMC already has hundreds of them, with a total cost of tens of billions of dollars. And if you want to use High-NA EUV in the 2nm and below process, then you need to invest more than 30 billion US dollars.
It can be said that the high cost of lithography machines is becoming a major pain point restricting the technological progress of the chip industry. For a leading enterprise like TSMC, this huge capital expenditure has undoubtedly seriously compressed its profit margins and investment capabilities.
Therefore, TSMC urgently needs to find a new response to continue to promote the downsizing of the chip manufacturing process without increasing too much cost. And its recent announcement of this decision is undoubtedly an important manifestation of this strategy.
TSMC's "table-lifting" move
Faced with the continuous upgrading of lithography machines, TSMC finally made a decision that shocked the industry: in the future production of chips of 2nm and below, it does not plan to use expensive High-NA EUV, but continues to use existing old EUV lithography machines.
This is undoubtedly "flipping the table". Because according to industry practice, every breakthrough in the chip manufacturing process means a comprehensive update of the corresponding lithography machine. For example, in the 7nm process, TSMC must switch from DUV to EUV. And now, it has announced that it will continue to use old EUV devices in 2nm and beyond.
This is undoubtedly a subversive decision for the entire chip mucus industry. Because in the past, ASML, as the sole supplier of EUV lithography machines, has been a key force driving the technological progress of the entire industry by continuously introducing more advanced equipment.
And now, TSMC's approach is undoubtedly challenging this pattern. Not only does it not require the purchase of expensive High-NA EUVs, but even ASML's predecessor EUV models, which it intends to continue to keep and utilize. This is undoubtedly a major blow to ASML.
Not only that, TSMC's move may also affect the entire chip industry chain. In the past, process downsizing often meant a comprehensive upgrade of the corresponding lithography machine, which brought new market opportunities to upstream and downstream manufacturers. But now, if TSMC can achieve 2nm and below processes on existing EUV models, then this market landscape is bound to change.
It can be said that TSMC's move to "flip the table" has undoubtedly had a far-reaching impact on the entire chip industry. It not only challenges the technological path of the industry, but also may reshape the ecological pattern of the entire industrial chain. What does this mean exactly? Let's find out.
TSMC's path to technological innovation
So, what kind of innovative technology does TSMC rely on to achieve 2nm and below processes on existing old EUV lithography machines? What are the unique technical ideas behind this?
First of all, we need to understand the basic working principle of EUV lithography machine. Unlike traditional DUV lithography technology, EUV lithography uses EUV light with shorter wavelengths, which allows it to achieve finer lithography to support smaller process nodes.
However, there are some technical bottlenecks in EUV lithography itself. For example, higher requirements are put forward for light source power, optical design, mask manufacturing, etc., which makes the cost of EUV equipment rise significantly. Of course, these problems are not insurmountable, and as the technology continues to improve, the performance and cost performance of EUV lithography machines continue to be optimized.
TSMC's key innovation lies in fully tapping the technical potential of existing EUV equipment.
First of all, TSMC has deeply studied the key parameters of the EUV lithography machine, including optical design, light source power, mask manufacturing, and so on. Through the fine optimization of these details, TSMC has found a technical path to achieve the process of 2nm and below on existing EUV models.
For example, in terms of optical design, TSMC engineers have targeted improvements to the imaging system of the EUV lithography machine to improve its resolution and imaging quality. At the same time, they also made innovations in light source power management and mask manufacturing processes, which effectively improved the overall performance of the EUV lithography machine.
At the same time, TSMC also focuses on using software technology to improve the accuracy and efficiency of lithography. For example, advanced computer simulation technology is used to digitally model and optimize the whole process of lithography to eliminate various process errors to the greatest extent.
It can be said that TSMC's innovative approach of "combining software and hardware" gives full play to the technical potential of existing EUV equipment, thereby achieving a breakthrough in the process of 2nm and below without the need to purchase expensive High-NA EUV. This not only saves a lot of capital investment, but also brings new possibilities to TSMC's future technology path.
Profound impact on the entire industry
It is worth noting that TSMC's "table-lifting" move not only breaks the industry convention, but may also have a far-reaching impact on the entire chip industry chain.
First of all, this decision will undoubtedly be a major blow to ASML. As the sole supplier of EUV lithography machines, ASML has long been a key force in driving technological advancement across the industry. But now, TSMC can actually implement the 2nm process on the old model, which undoubtedly weakens ASML's monopoly in technology upgrades.
This is undoubtedly a big challenge for ASML's business development. In the past, ASML has been able to bring stable orders and revenues to itself by continuously introducing more advanced lithography models. However, if a leading enterprise like TSMC no longer needs to update lithography equipment frequently, ASML's market space will inevitably be compressed to a certain extent.
At the same time, TSMC's move may also reshape the pattern of the entire chip industry chain. In the past, as the process continued to shrink, upstream and downstream manufacturers needed to continuously upgrade equipment and process adjustments, which brought new market opportunities to them. But now, if TSMC can implement the 2nm process on existing EUV models, then this landscape is bound to change.
Some equipment manufacturers and upstream material suppliers may lose their original market opportunities because of TSMC's decision. Correspondingly, TSMC may also further consolidate its leading position in the industrial chain, so as to gain more voice and dominance.
In general, TSMC's innovative decision has undoubtedly triggered in-depth thinking in the entire chip industry. It not only challenges the technological path of the industry, but also may reshape the competitive landscape of the entire industry chain. This innovative spirit is undoubtedly worthy of learning and reference from the industry.
epilogue
For TSMC, its decision to "flip the table" undoubtedly reflects a rare courage and strategic vision. In the context of the rising cost of lithography machines, it actively explored the potential of existing technologies and found innovative ways to break through bottlenecks, and finally achieved a breakthrough in the manufacture of chips of 2nm and below on old EUV equipment.
This spirit of innovation has not only won TSMC's own competitive advantages in technology and cost, but also has a profound impact on the entire chip industry. It not only challenges the technological path of the industry, but also may reshape the competitive landscape of the industrial chain and bring new possibilities for future technological development.
We believe that under the leadership of TSMC, the entire chip industry will set off a new technological revolution. Whether it is an innovative breakthrough in lithography machines or a reshaping of the ecology of the industrial chain, TSMC's "table-lifting" move will surely become a memorable milestone in the history of technological development.