Dutch Tech Giant ASML Eyes Major Expansion Into AI Chip Manufacturing Equipment

ASML Holding, the world's only maker of extreme ultraviolet chipmaking equipment, is planning significant expansion beyond its current technology to capture more of the booming artificial intelligence chip market. The Dutch company aims to develop new tools for advanced chip packaging and connecting multiple specialized processors that power AI systems.

A major Dutch technology company that plays a crucial role in manufacturing the world’s most advanced computer chips is setting its sights on significant expansion to capitalize on the artificial intelligence boom.

ASML Holding, currently the sole producer of extreme ultraviolet (EUV) machinery essential for creating cutting-edge AI processors, has revealed ambitious plans to broaden its equipment offerings, according to a senior company official speaking with Reuters.

The Netherlands-based corporation has spent over ten years developing its EUV technology, investing billions of dollars in systems that companies like Taiwan Semiconductor Manufacturing Co and Intel depend on for producing the most sophisticated AI chips available today. While continuing to advance its EUV capabilities with next-generation products in development, ASML is now looking to diversify its portfolio.

The company’s strategy involves entering the advanced packaging market, which focuses on creating tools that bond and link multiple specialized processors together – a critical component in AI chip construction and the high-performance memory systems that support them. This expansion will also incorporate artificial intelligence into both new ventures and existing operations.

“We look, not just for the next five years, we look at the next 10, maybe 15 years,” Chief Technology Officer Marco Pieters explained to Reuters. “(We look at) what are potential directions the industry could take, and what would it require in terms of packaging, bonding, etc.?”

ASML’s current EUV systems perform lithography, utilizing light to create intricate patterns on silicon wafers during chip production. The company is also investigating whether it can increase the maximum chip size beyond its present limitation – approximately the dimensions of a postage stamp – which currently restricts processing speed.

The technology leadership recently underwent changes when Pieters was elevated to CTO in October, taking over from Martin van den Brink, who led the technology division for nearly four decades. ASML restructured its technology operations in January to emphasize engineering positions over administrative roles.

Wall Street has high expectations for Pieters and CEO Christophe Fouquet, who assumed his position in 2024. The company’s stock reflects investor confidence in its EUV market dominance, trading at approximately 40 times forward earnings compared to Nvidia’s 22 times earnings ratio. ASML’s market capitalization of $560 billion has seen shares climb more than 30% this year.

The corporation is accelerating efforts to manufacture chip packaging machinery and developing equipment for next-generation AI processor production.

“We’re actually researching that – to what extent can we participate in it, or what we can add to that part of the business,” Pieters stated.

Drawing from his software development background at ASML, Pieters noted that as the company’s equipment becomes faster, engineers will implement AI to accelerate machine control software and enhance chip inspection processes during manufacturing.

The chip design landscape has transformed dramatically in recent years. Previously, companies like Nvidia and Advanced Micro Devices created processors that were essentially flat, resembling single-level structures. Today’s chips increasingly resemble multi-story buildings with various levels connected through microscopic pathways.

Due to size constraints, combining chips in vertical stacks or horizontal arrangements allows designers to boost processing speeds for the complex computations needed to build large AI models or operate chatbots like OpenAI’s ChatGPT.

The precision and complexity required for constructing these multi-layered chips has transformed packaging from a low-profit, high-volume industry into a more profitable manufacturing segment for companies like ASML. TSMC has employed advanced packaging techniques to build Nvidia’s most sophisticated AI processors.

“But we also see more of that advanced packaging is coming to the front end,” Pieters observed, referring to TSMC and similar companies’ activities. “Accuracy is becoming more and more important.”

After analyzing chip manufacturers’ roadmaps – including memory producers like SK Hynix – Pieters recognized the growing demand for additional machinery to help companies produce vertically stacked chips and similar configurations.

In the previous year, ASML unveiled a scanning device called the XT:260, specifically designed to assist in manufacturing advanced memory chips for AI applications and AI processors themselves. Company engineers are currently exploring additional equipment possibilities, according to Pieters.

“One of the things I’m doing is also looking at what could be a product portfolio in that direction,” Pieters revealed.

As AI chips continue growing in size, the company is investigating supplementary scanner systems and lithography equipment to enable even larger chip production.

The scanning technology leverages expertise in optics and specialized knowledge about precise silicon wafer handling, giving ASML advantages in developing future equipment, Pieters explained.

“It will co-exist next to what we’ve been doing for the last 40 years,” he concluded.