Focus on Computational Lithography’s ‘new track.’
Moores’s Law predicts the development trajectory of chip technology. Despite the appearance of ‘steady’ development at process nodes, the photolithography technology required for chip manufacturing has undergone explosive innovation. In the 21st century, the emergence of immersion lithography broke the 193nm wavelength limit. As line widths continue to shrink, the complexity of lithography processes has increased once again. Relying solely on updates to semiconductor equipment no longer meets the demands of chip manufacturing. Photolithography technology is once again at a crossroads, needing to choose its direction.
The software can achieve more precise photolithography. Computational lithography (also known as the OPC field) has played a significant role in chip manufacturing since the 90nm node. Meanwhile, as advanced processes continue to advance, the difficulty of developing semiconductor equipment for chip manufacturing and the energy consumption and R&D investment required has steadily increased. In this context, the development of computational lithography has garnered industry attention and become a critical technology for overcoming the bottlenecks in advanced process development and entering the post-Moore era.
01
The Importance of Advancing Computational Lithography
As chip manufacturing continues to shrink linewidth dimensions, the Optical Proximity Effect (OPE) becomes increasingly pronounced, significantly impacting the yield of the exposure and development process. To counteract these errors and ensure that the resulting images meet design requirements, Optical Proximity Correction (OPC) must be applied to the patterns on the mask. Computational lithography can perform targeted OPC on each pattern, optimizing mask shapes and sizes based on predicted distortions, thus achieving high-quality chip pattern imaging on wafers.
Computational lithography has become an indispensable part of the chip manufacturing process. It can be said that without computational lithography, even with lithography machines, chip production would not be possible. Without computational lithography software, all chip manufacturers would lose the ability to transform chip designs into actual chip products.
In the context of the semiconductor industry’s development, global competition in semiconductor manufacturing revolves around chip production. Computational lithography substantially enhances chip manufacturing efficiency, making it a necessity in the field. As a key technology in high-end chip manufacturing, investing in computational lithography has become an industry consensus.
From a technological perspective, as chip linewidths continue to shrink, various optical effects, system errors, and process condition deviations become increasingly intricate. Computational lithography addresses nanometer-level mask repair, chip design, manufacturing defect detection, and correction through algorithmic modeling, simulation calculations, data analysis, and result optimization. It is the primary method for improving imaging performance in lithography systems. With the ongoing escalation of problems and complexities in chip manufacturing processes, computational lithography’s workload continues to grow, emphasizing the industry’s focus on advancing computational lithography technology.
From a market demand standpoint, as a crucial branch of the EDA field, computational lithography’s market share in the semiconductor sector may not be high currently. However, as major semiconductor players globally compete to enhance chip manufacturing capabilities, the demand for computational lithography technology is rapidly increasing. Particularly in today’s emphasis on localization, the global semiconductor arms race will directly drive the rapid development of computational lithography technology domestically. To secure a foothold in the semiconductor manufacturing industry, substantial development of computational lithography technology and the nurturing of relevant talent is crucial for the sustainable and healthy growth of the domestic semiconductor industry.
02
Overview of the Computational Lithography Market
At the beginning of this year, NVIDIA introduced the cuLitho solution with GPU support, focusing on advancing computational lithography to enhance chip production efficiency from a high-performance computing perspective. This move has generated significant industry interest and attention towards computational lithography, making it a highly watched segment. In terms of market size, the global computational lithography software market was estimated to be around $1 billion in 2022. Over the next five years, this market is expected to grow to over $2 billion.
Computational lithography is an automated process that optimizes semiconductor fabrication using an ‘algorithm + data’ approach. It is essentially an EDA simulation technology within the semiconductor manufacturing process. Therefore, traditional EDA companies like Siemens EDA (formerly Mentor Graphics) and Synopsys are participants in the computational lithography market.
Apart from EDA vendors, a hardware-centric company, ASML, a giant in photolithography machines, is also a significant player in the global computational lithography market. Initially, computational lithography was known as part of ASML’s ‘iron triangle’ comprehensive lithography solution. While ASML continued to develop core technologies in the computational lithography field, its strength was somewhat overshadowed by its photolithography machine business. It wasn’t until the computational lithography market gained prominence that ASML’s capabilities in this area became more evident. ASML’s dual focus on software and hardware development has solidified its leading position in the field of lithography.
The increased investment by top-tier companies in computational lithography technology has drawn continuous industry attention to this field. In recent years, domestic companies in various countries have also begun to catch up with computational lithography technology, with some local players gradually emerging as significant contributors.
03
Challenges in the Development of Computational Lithography
The global computational lithography software market exhibits a high level of concentration and significant technical barriers. International vendors typically possess years of technical expertise, extensive industry experience, and a solid customer base, which gives them a considerable advantage in terms of customer relationships and trust. For domestic semiconductor industries, especially local computational lithography companies, overcoming these challenges is essential.
Computational lithography, as an EDA tool for chip manufacturing, presents higher development complexities compared to other EDA software categories. It falls into a technology, capital, and talent-intensive field. The lengthy research and development cycle results in critical factors restraining industry growth, including talent demand and capital consumption.
Regarding research and development investment, computational lithography is a capital-intensive business with long development cycles. Continuous and substantial R&D investment not only fosters innovative technologies but also serves as a source of competitiveness for companies. However, due to the limited overall scale of the computational lithography market and the considerable gap between domestic chip manufacturing capabilities and leading international companies, local enterprises may hesitate to enter the market. Therefore, the domestic development of computational lithography requires sustained attention and policy support.
Talent is the driving force behind innovative development in computational lithography technology. When domestically developing computational lithography technology, relevant policies should prioritize the cultivation of talent in this field. The computational lithography industry primarily requires hybrid professionals with expertise in at least two of the following categories: software, chip design, physics, and mathematics/microelectronics. For example, product development roles primarily involve writing software code, often using C++ language, focusing on the accuracy of results and algorithm speed. This requires professionals with a deep understanding of software, physics, and mathematics/microelectronics. Once products are developed, application engineers with knowledge of chip design and manufacturing are needed to promote software solutions to chip manufacturers. Qualified and exceptional talent in the computational lithography industry is relatively scarce worldwide. While the industry has increased efforts to train computational lithography professionals in recent years, there are challenges such as high demand, a shortage of specialized talent, and long training periods. Talent scarcity is likely to remain a long-term challenge.
04
How to Promote the Development of Computational Lithography?
The development of computational lithography in China is still in its early stages, and it is crucial to promote the healthy growth of the computational lithography industry rationally.
- Seize the Opportunity in the Chinese Market: The semiconductor industry is of great strategic importance, and the domestic demand for chips in China is enormous. Moreover, the government’s investments in the semiconductor manufacturing sector are expected to continue to grow. These factors will drive rapid growth in the computational lithography market. For leading foreign companies, the Chinese market presents unlimited development potential. It is essential to develop tailored market strategies that consider local conditions. For domestic enterprises, localization can be a significant opportunity. Strengthening technology research and development, collaborating closely with manufacturing companies, and enhancing the industry’s supply chain will all be essential for rapid development.
- Promote Computational Lithography Development Rationally: Excessive capitalization can disrupt the normal development of the computational lithography market. Computational lithography is not a field where quick results can be achieved; it requires long-term technical accumulation and investment. Companies must be willing to invest both time and substantial resources in order to build long-term competitiveness.
- Invest Significantly in the Ongoing Training of Local Professionals: Computational lithography software application requires close integration with engineering implementation. Local talent development should focus on serving the growth of the domestic chip manufacturing industry and promoting the deep cultivation of computational lithography technology in China. ASML’s commitment to nurturing computational lithography talent serves as a valuable example. Since establishing its computational lithography research and development center in China in 2004, the center has become ASML’s largest professional software research and development base in Asia. Over two decades, ASML has trained computational lithography experts who have become essential for the company’s stable long-term development in China.
- Collaborate with Upstream and Downstream Partners to Break Down Ecosystem Barriers: Computational lithography, as a form of EDA for manufacturing, presents significant ecosystem barriers. Solving all problems solely through one EDA company is not feasible. It is necessary to establish collaborations across the key segments of the supply chain, working closely with chip manufacturers to build an ecosystem and address practical application challenges.
05
Words in the End
The evolution of semiconductor fabrication processes has encountered bottlenecks, making the emergence of new technologies more attention-grabbing for the industry. With the high and persistent costs of hardware equipment in chip manufacturing, software is poised to become a new breakthrough for process advancements. Computational lithography, as a ‘new track’ driving the development of chip manufacturing, is certain to garner significant attention within the domestic industry.
Recommended Reading:
- IP Merge in Semiconductor Design: Key Overview
- FPGA Chips: Why They Aren’t Suited for Algorithms
- Green Web Transformation: Save and Earn Money
- Why Python Is the Top Choice for Artificial Intelligence
- Is Access Database Still Necessary in Office Today?
- Keil MDK Bin Format Guide – Everything You Should Know
Disclaimer: This article is created by the original author (ICWISE). The content of the article represents their personal opinions. Our reposting is for sharing and discussion purposes only and does not imply our endorsement or agreement. If you have any objections, please contact us through the provided channels.
