TEMC: Building a Resilient Semiconductor Specialty Gas Supply Chain for Advanced Memory and Logic Chips

In the global market, many experts believe Korea is in a particularly advantageous position. Not only does the country have clear strengths such as speed of execution and strong customization capabilities but Korea can also benefit from the ongoing realignment of global supply chains. Do you agree with this positive sentiment? And for your company in particular, what opportunities do you see today in the global market?

I do agree with that perspective. Across the semiconductor industry, customers today place enormous emphasis on securing a stable, resilient supply chain. This has been heightened by recent trade disputes, particularly those involving the United States and China.

Historically, companies prioritized cost efficiency when evaluating materials supply. Stability was important, but it was secondary. That mindset has shifted. Today, supply stability outweighs cost considerations, especially given our industry’s heavy dependency on China for certain critical materials.

To mitigate this dependency, companies are actively seeking diversified and reliable supply routes that can support the entire semiconductor ecosystem. This dynamic has created meaningful global opportunities for Korean suppliers like TEMC companies that are agile, technologically capable, and well-positioned to help customers build more resilient supply networks.

Since 2024, the industry has rebounded strongly, driven by new applications particularly AI and by the opening of new fabs worldwide. We see large-scale projects in the U.S. by TSMC, Samsung, and Intel, and similarly in Korea, from the Yongin cluster to SK Hynix’s new M15X fab. However, the recovery has not been uniformed across all applications. In memory especially NAND the rebound has been slower due to weaker consumer demand, high inventories, cost pressures on manufacturers, and environmental considerations that directly affect the industrial gas segment. What are your short-term expectations for the NAND materials market? And looking toward 2030, how do you anticipate the sector will grow?

In the short term, we expect the NAND market to recover, but that recovery is limited by ongoing inventory adjustments and reduced production levels. Demand will return, but it is still stabilizing.

Looking further ahead, when the semiconductor market enters a full upcycle driven significantly by AI memory inevitably follows. While CPUs and GPUs may lead the initial wave, DRAM and NAND are essential enablers of data-intensive AI workloads. Historically, memory volumes outpace logic when growth accelerates, and we expect that pattern to continue.

The downturn that began three years ago has now reached its turning point. During that period, the key question was how quickly AI technology would mature. Today, the question is no longer when AI will grow, but how far it will expand. At a forum earlier today, a KAIST professor emphasized that AI is still at its very beginning and may require computing power to grow by a factor of one thousand. If that trajectory unfolds, advanced memory HBM, DRAM, NAND, and new 3D architectures will be essential.

From that standpoint, the long-term outlook for memory materials is very bright.

Turning to rare gases: this segment, like the semiconductor industry as a whole, is highly cyclical, especially since gas prices are openly traded and prone to major fluctuations. For example, during the Russia-Ukraine war, prices skyrocketed, and in 2022 companies like yours recorded record-high revenues as a result. For a company like TEMC, there is a clear need to reduce exposure to such volatility and mitigate cyclicality by diversifying your offerings. Could you walk us through your diversification strategy how you plan to reduce reliance on cyclical materials and broaden your portfolio?

TEMC is one of the Korean companies that has achieved the highest degree of localization in specialty gases. Our long-term objective is to establish safe localization, meaning that we develop the capability to produce domestically and then export internationally.

However, because Korea lacks natural reserves of many rare materials, our next strategic priority is recycling and recovery.

Take germanium as an example: In China, vast quantities of ore are processed to extract a small amount of high-purity germanium used in semiconductors. Only about 10% is utilized, while roughly 90% is discarded. If we can recover that unused germanium not by mining, but through recycling streams it would be a breakthrough for the materials industry.

Rare materials are expensive, difficult to produce, and fundamental to semiconductor manufacturing. Recycling them is essential not only for economic efficiency but also for global supply-chain stability.

Additionally, customers themselves have experienced supply-chain vulnerabilities in recent years. Their demand for stable, renewable sources of rare gases and materials has grown significantly. TEMC intends to meet that demand by developing advanced recovery technologies.

This is particularly interesting given that dependency on rare materials has become a geopolitical concern. China controls much of the world’s supply, and governments globally worry about over-reliance. Your neon-recovery partnership with SK Hynix achieving more than 70% recovery has been widely recognized. Given that this is a global problem, what is the international potential of your recovery technology? Do you plan to deploy it in the U.S. or Europe?

Yes, the international potential is significant. Recycling is not limited to Korea; it is directly applicable to our operations overseas. When shortages occur, they typically occur globally, not locally. Conversely, when even one company succeeds in recycling a scarce material, the entire industry benefits.

Recycling rare materials supports consistent supply, reduces price volatility, and decreases dependence on any one country. It strengthens not only our own operations but the global semiconductor value chain.

TEMC portfolio

Having followed the semiconductor materials field for many years previously working with companies like Nippon Sanso and Sumitomo Seika I’ve been impressed by how TEMC broke what was essentially a long-standing oligopoly. Given the dominance of Japanese and European suppliers historically, what do you believe enabled TEMC to succeed where others did not?

I believe three key factors contributed to our success.

First: timing. TEMC entered the market precisely as NAND flash technology transitioned from 2D to 3D. This shift fundamentally changed the required gas materials. Typically, gas requirements do not change often, but at that moment they did and we responded quickly. Our agility at that critical juncture determined our success.

Second: environment specifically talent and technology. The “T” in TEMC stands for Talent. We prioritize attracting capable people and cultivating strong technological expertise. But even with talent and technology, success requires customers who support localization. We were fortunate to have such partners.

Third: we had the right products at the right time, and strong customer relationships. Smaller companies rarely outperform large global suppliers because of capital constraints and risk aversion among clients. However, in our case, the combination of timing, environment, and customer trust allowed us not just to compete, but to succeed.

TEMC’s growth has been anchored in the memory segment, particularly NAND. You have long-standing relationships with SK Hynix and Samsung, and now also supply Kioxia and Micron. Given that materials such as neon, IMP gases, and ArF excimer gases can also serve logic applications, how do you plan to grow your customer base? Will you continue prioritizing memory, or will you expand into logic-oriented foundries?

The boundary between memory and logic is becoming increasingly blurred. TSMC and Samsung are developing 1.5nm and 2nm processes. DRAM is moving below 10nm, and NAND continues raising its 3D stack layers. As technology converges, material requirements begin to overlap.

We are actively benchmarking the materials used in advanced logic processes, while also focusing on next-generation memory technologies such as 3D DRAM. Our aim is not to choose between memory and logic, but to position ourselves around the most advanced materials required across both domains.

TEMC CNS

In 2023, you acquired Oceanbridge now TEMC CNS giving you access to precursor systems and even equipment for secondary-battery manufacturing. Nearly two years later, what synergies have emerged from this acquisition, and how has it strengthened the TEMC group?

TEMC’s core business is specialty gases. TEMC CNS complements us with expertise in precursors and chemical supply systems. This significantly broadens our materials portfolio.

Another important aspect is that the semiconductor industry is driven heavily by equipment investment cycles. Through TEMC CNS and its affiliates, including JEIL Eng, we now have deeper insight into equipment trends, which helps us anticipate future materials demand.

We are also exploring new equipment technologies such as x-ray inspection systems for testing material quality. Although still in early stages, this diversification enhances our understanding of the industry and strengthens our strategic position.

Are you planning additional acquisitions in the equipment segment in the coming years?

We are still at an early stage, but entering the equipment sector provides valuable insight into technological shifts. Every major innovation in semiconductors requires new equipment, which in turn requires new materials. Strengthening our position in equipment allows us to better anticipate and prepare for future materials demand.

I guess that is also why you are investing in equipment to diversify your offering and reduce cyclicality. Is that correct?

Exactly. Diversifying into equipment enables us to reduce exposure to cyclical fluctuations and understand future material trends more clearly. It supports our long-term strategy of stability and sustainable growth.

Global Network

Beyond Korea, TEMC now maintains bases in Europe, including Hungary, as well as in the United States. What are your expectations for the American and European markets? And how important are these overseas bases for your long-term expansion?

Domestically, TEMC has established a strong position in specialty gases. We supply not only Samsung and SK Hynix but also global leaders such as TSMC, Kioxia, Micron, Intel, and customers in Singapore and the United States.

However, one area we must strengthen is global networking. Our products require delivery, storage, usage, and recovery creating cyclical risks that must be supported by a reliable international infrastructure.

That is why we are considering manufacturing sites in regions like Singapore and Japan. Multinational major gas players maintain their global leadership partly because of their strong international networks. For TEMC, building that network is essential.

I suppose that is also why you are investing in equipment to broaden your offering and reduce cyclicality.

Yes, diversification supports our goal of stabilizing our business model and preparing for long-term growth.

If I may ask an investor-relations question: When TEMC IPO-ed in 2023, your listing price was about 15,000 won. Within the first year, the stock rose to around 29,000 won, but over the last year it has gradually declined to about 7,000 won today. This decline contrasts with the positive momentum of your business your collaborations with POSCO and SK Hynix, your neon-recovery success, and your global expansion within an industry projected to grow strongly from 2025 to 2030. Where do you think the disconnect lies between market valuation and the company’s actual performance and long-term potential?

It’s a question I reflect on as well. TEMC consistently generates around 50 billion won in annual EBITDA (Earnings before interest, taxes, depreciation and amortization). Even during industry downturns when many equipment and component companies see revenue fall dramatically, we remain profitable. We do not experience extreme volatility in performance.

One challenge is the absence of peer companies in Korea. Globally, multinational major gas players are recognized and traded at high multiples. China has several comparable firms as well. But Korea lacks peer benchmarks in specialty gases, which affects investor perception.

Another factor is growth perception. For example, the industry has recently seen an explosive expansion of AI from multi-modal models to agent-based AI and physical AI, each of which is causing an exponential increase in computing demand. This growth requires advanced memory such as HBM, yet global HBM supply remains insufficient.

From that standpoint, the memory industry is only at its beginning, and TEMC plays a strategically important role within it. Our growth potential is not theoretical it is anchored in the structural evolution of semiconductors.

In my view, the gap between valuation and potential reflects market communication rather than business fundamentals.

Interested in learning more? Click here: https://temc.co.kr/