Marketdash
Back to News

InnoScience Wins a Round in Patent Fight, But the Real Battle Is Just Beginning

MarketDash Editorial Team
4 hours ago
The Chinese GaN semiconductor maker got mostly good news from U.S. trade regulators in its patent dispute with Infineon, sending shares up nearly 10%. But while legal victories are nice, the company still faces a more fundamental problem: everyone and their cousin is piling into the gallium nitride market.

Nearly two years into a legal battle with industry heavyweight Infineon, InnoScience (Suzhou) Technology Holding Co. Ltd. (2577.HK) got some welcome news last week. The U.S. International Trade Commission issued a ruling that the Chinese power semiconductor maker says was largely in its favor, potentially clearing a path for future expansion into American markets.

The stakes in this dispute, which kicked off in March 2024, weren't trivial. While InnoScience currently does most of its business in China, an unfavorable ruling could have blocked it from selling certain products in the U.S. down the road. That's the kind of thing that keeps management teams up at night, especially when you're trying to go global.

According to InnoScience's announcement to the Hong Kong Stock Exchange, the ITC ruled partly in its favor on one of two patents in the case, and the company says its products don't infringe on the other patent either. The company described the ruling as clarifying "the status of InnoScience's relevant intellectual property" and clearing "the way for its future global development."

Infineon, unsurprisingly, had a different take. In its own statement, the German chipmaker pointed out that the ITC actually found in its favor regarding one of the patents. More importantly, the final decision isn't expected until April 2, 2026, and if that decision confirms the initial ruling, it could still lead to a ban on related InnoScience products entering the U.S.

So who's right? Investors seem to think InnoScience came out ahead. The stock dipped 1.5% on the day of the announcement last Wednesday, but quickly bounced back and now trades nearly 10% higher than pre-announcement levels. That's a pretty clear vote of confidence.

The GaN Revolution

To understand what's at stake here, you need to know a bit about gallium nitride, or GaN. Since its founding in 2017, InnoScience has been championing this niche but increasingly important technology in the power semiconductor world. GaN offers some real advantages over traditional silicon-based semiconductors: higher efficiency, faster switching speeds, and better performance in demanding applications.

InnoScience wasn't just talking about GaN, it was doing something about it. The company became the first to mass-produce 8-inch GaN wafers, giving it a technological edge in what's shaping up to be a rapidly growing market. According to Bank of America, the GaN power semiconductor market is expected to explode from 2.7 billion yuan ($382 million) in 2024 to 19.2 billion yuan in 2028. That's 63% annual growth, which is the kind of trajectory that gets investors excited.

Right now, InnoScience sits at the top of the heap with 30% market share in 2024, nearly double second-place Navitas (NVTS) at 17%, per Bank of America's analysis.

The IDM Advantage

One reason InnoScience has pulled ahead is its integrated device manufacturer (IDM) model. Unlike fabless competitors that outsource manufacturing to third parties, InnoScience handles everything from design to production to marketing. This vertical integration gives the company more control over its supply chain and product quality, though it comes with a major downside: massive upfront capital costs for building production facilities.

But here's where the IDM model is really paying off. Navitas, which uses the fabless approach, is currently scrambling after production partner TSMC announced plans to exit the GaN manufacturing business in 2027. InnoScience doesn't have that problem.

The IDM model also diversifies InnoScience's revenue streams. The company doesn't just make end products like GaN chips and modules, it also produces wafers, the raw material from which chips are made. Two recent supply deals highlight this advantage: one announced last week with U.S.-based ON Semiconductor, and an earlier agreement with Europe's STMicroelectronics. Both deals will see InnoScience supplying GaN wafers to these major players, helping diversify the company's geographic revenue mix beyond its current heavy China concentration.

Morgan Stanley, in an October note, projected that InnoScience's revenue would grow even faster than the overall GaN market, expanding 66% annually through 2027. The investment bank credited this outlook largely to the company's technological leadership and IDM advantages.

The Numbers Tell a Story

That growth is already showing up in InnoScience's financial results. In the first half of 2025, revenue jumped 43.4% to 553.4 million yuan ($71.1 million). Even more significantly, the company achieved its first ever positive gross profit margin of 6.8%. For a young company in capital-intensive semiconductor manufacturing, crossing that threshold matters.

The company still posted a net loss of 429 million yuan for the period, though that's an improvement from the 488 million yuan loss in the same period a year earlier. Progress, not perfection.

Another major win came in July when global AI chip leader Nvidia announced it had selected InnoScience to supply 800V DC power solutions for its Rubin Ultra GPU, slated for commercialization in 2027. This represents an important breakthrough into the data center segment, one of the fastest-growing areas for GaN power semiconductors. When Nvidia picks you as a supplier, people notice.

Investors certainly noticed. InnoScience's shares climbed steadily after its Hong Kong listing last December, more than tripling from their IPO price of HK$30.86 to hit a peak of nearly HK$100 in early September. The stock has since moderated to HK$80.85 as of last Friday's close, but that's still a pretty impressive run.

The Competition Problem

Here's where things get tricky. InnoScience is aggressively expanding capacity, targeting 60,000 wafers per month by 2027, up from 13,000 in the first half of this year. That's a necessary move if you want to capture growth in a booming market. But InnoScience isn't the only one building out capacity. Not even close.

Morgan Stanley forecasts that global GaN capacity will surge from 100,000 wafers per month in 2024 to 345,000 6-inch wafer equivalents by 2029. That's more than tripling in five years. If you know anything about the semiconductor industry, you know what comes next: overcapacity, pricing pressure, and squeezed margins.

The chip sector is famous for its boom-bust cycles driven by exactly this kind of overbuilding. Everyone sees the growth potential, everyone builds capacity, and suddenly there's too much supply chasing demand. Morgan Stanley expects overall industry utilization to improve but remain below 60% over the next five years, which suggests persistent pricing and profitability pressures across the entire GaN market.

As a result, the investment bank expects InnoScience's gross margin to improve only at a moderate pace in 2026 and 2027, with the company continuing to lose money through at least next year. That's the reality of being in a capital-intensive business where everyone's racing to build capacity at the same time.

The Bottom Line

So where does that leave InnoScience? The ITC ruling is genuinely good news, removing a legal cloud that had been hanging over the company and potentially opening doors in the U.S. market. The company's technological leadership is real, its IDM model provides meaningful advantages, and partnerships with players like Nvidia and ON Semiconductor validate its position in the market.

But let's be clear about what this ruling doesn't do: it doesn't address the fundamental competitive dynamics facing the company. InnoScience may be winning the patent battle, but the bigger war is about capacity buildouts and margin pressure. The GaN market is growing fast, but so is everyone else's ability to serve it. The path from technological leadership to sustained profitability is rarely a straight line, especially when the entire industry is building capacity simultaneously.

InnoScience's potential to become a dominant force in the GaN market is real, but it's far from guaranteed. Investors betting on the company are essentially wagering that its technological edge and integrated manufacturing model will allow it to weather the coming period of industry overcapacity and emerge as one of the winners. That's a reasonable bet, but it's definitely still a bet.

Back to News

InnoScience Wins a Round in Patent Fight, But the Real Battle Is Just Beginning

MarketDash Editorial Team
4 hours ago
The Chinese GaN semiconductor maker got mostly good news from U.S. trade regulators in its patent dispute with Infineon, sending shares up nearly 10%. But while legal victories are nice, the company still faces a more fundamental problem: everyone and their cousin is piling into the gallium nitride market.

Nearly two years into a legal battle with industry heavyweight Infineon, InnoScience (Suzhou) Technology Holding Co. Ltd. (2577.HK) got some welcome news last week. The U.S. International Trade Commission issued a ruling that the Chinese power semiconductor maker says was largely in its favor, potentially clearing a path for future expansion into American markets.

The stakes in this dispute, which kicked off in March 2024, weren't trivial. While InnoScience currently does most of its business in China, an unfavorable ruling could have blocked it from selling certain products in the U.S. down the road. That's the kind of thing that keeps management teams up at night, especially when you're trying to go global.

According to InnoScience's announcement to the Hong Kong Stock Exchange, the ITC ruled partly in its favor on one of two patents in the case, and the company says its products don't infringe on the other patent either. The company described the ruling as clarifying "the status of InnoScience's relevant intellectual property" and clearing "the way for its future global development."

Infineon, unsurprisingly, had a different take. In its own statement, the German chipmaker pointed out that the ITC actually found in its favor regarding one of the patents. More importantly, the final decision isn't expected until April 2, 2026, and if that decision confirms the initial ruling, it could still lead to a ban on related InnoScience products entering the U.S.

So who's right? Investors seem to think InnoScience came out ahead. The stock dipped 1.5% on the day of the announcement last Wednesday, but quickly bounced back and now trades nearly 10% higher than pre-announcement levels. That's a pretty clear vote of confidence.

The GaN Revolution

To understand what's at stake here, you need to know a bit about gallium nitride, or GaN. Since its founding in 2017, InnoScience has been championing this niche but increasingly important technology in the power semiconductor world. GaN offers some real advantages over traditional silicon-based semiconductors: higher efficiency, faster switching speeds, and better performance in demanding applications.

InnoScience wasn't just talking about GaN, it was doing something about it. The company became the first to mass-produce 8-inch GaN wafers, giving it a technological edge in what's shaping up to be a rapidly growing market. According to Bank of America, the GaN power semiconductor market is expected to explode from 2.7 billion yuan ($382 million) in 2024 to 19.2 billion yuan in 2028. That's 63% annual growth, which is the kind of trajectory that gets investors excited.

Right now, InnoScience sits at the top of the heap with 30% market share in 2024, nearly double second-place Navitas (NVTS) at 17%, per Bank of America's analysis.

The IDM Advantage

One reason InnoScience has pulled ahead is its integrated device manufacturer (IDM) model. Unlike fabless competitors that outsource manufacturing to third parties, InnoScience handles everything from design to production to marketing. This vertical integration gives the company more control over its supply chain and product quality, though it comes with a major downside: massive upfront capital costs for building production facilities.

But here's where the IDM model is really paying off. Navitas, which uses the fabless approach, is currently scrambling after production partner TSMC announced plans to exit the GaN manufacturing business in 2027. InnoScience doesn't have that problem.

The IDM model also diversifies InnoScience's revenue streams. The company doesn't just make end products like GaN chips and modules, it also produces wafers, the raw material from which chips are made. Two recent supply deals highlight this advantage: one announced last week with U.S.-based ON Semiconductor, and an earlier agreement with Europe's STMicroelectronics. Both deals will see InnoScience supplying GaN wafers to these major players, helping diversify the company's geographic revenue mix beyond its current heavy China concentration.

Morgan Stanley, in an October note, projected that InnoScience's revenue would grow even faster than the overall GaN market, expanding 66% annually through 2027. The investment bank credited this outlook largely to the company's technological leadership and IDM advantages.

The Numbers Tell a Story

That growth is already showing up in InnoScience's financial results. In the first half of 2025, revenue jumped 43.4% to 553.4 million yuan ($71.1 million). Even more significantly, the company achieved its first ever positive gross profit margin of 6.8%. For a young company in capital-intensive semiconductor manufacturing, crossing that threshold matters.

The company still posted a net loss of 429 million yuan for the period, though that's an improvement from the 488 million yuan loss in the same period a year earlier. Progress, not perfection.

Another major win came in July when global AI chip leader Nvidia announced it had selected InnoScience to supply 800V DC power solutions for its Rubin Ultra GPU, slated for commercialization in 2027. This represents an important breakthrough into the data center segment, one of the fastest-growing areas for GaN power semiconductors. When Nvidia picks you as a supplier, people notice.

Investors certainly noticed. InnoScience's shares climbed steadily after its Hong Kong listing last December, more than tripling from their IPO price of HK$30.86 to hit a peak of nearly HK$100 in early September. The stock has since moderated to HK$80.85 as of last Friday's close, but that's still a pretty impressive run.

The Competition Problem

Here's where things get tricky. InnoScience is aggressively expanding capacity, targeting 60,000 wafers per month by 2027, up from 13,000 in the first half of this year. That's a necessary move if you want to capture growth in a booming market. But InnoScience isn't the only one building out capacity. Not even close.

Morgan Stanley forecasts that global GaN capacity will surge from 100,000 wafers per month in 2024 to 345,000 6-inch wafer equivalents by 2029. That's more than tripling in five years. If you know anything about the semiconductor industry, you know what comes next: overcapacity, pricing pressure, and squeezed margins.

The chip sector is famous for its boom-bust cycles driven by exactly this kind of overbuilding. Everyone sees the growth potential, everyone builds capacity, and suddenly there's too much supply chasing demand. Morgan Stanley expects overall industry utilization to improve but remain below 60% over the next five years, which suggests persistent pricing and profitability pressures across the entire GaN market.

As a result, the investment bank expects InnoScience's gross margin to improve only at a moderate pace in 2026 and 2027, with the company continuing to lose money through at least next year. That's the reality of being in a capital-intensive business where everyone's racing to build capacity at the same time.

The Bottom Line

So where does that leave InnoScience? The ITC ruling is genuinely good news, removing a legal cloud that had been hanging over the company and potentially opening doors in the U.S. market. The company's technological leadership is real, its IDM model provides meaningful advantages, and partnerships with players like Nvidia and ON Semiconductor validate its position in the market.

But let's be clear about what this ruling doesn't do: it doesn't address the fundamental competitive dynamics facing the company. InnoScience may be winning the patent battle, but the bigger war is about capacity buildouts and margin pressure. The GaN market is growing fast, but so is everyone else's ability to serve it. The path from technological leadership to sustained profitability is rarely a straight line, especially when the entire industry is building capacity simultaneously.

InnoScience's potential to become a dominant force in the GaN market is real, but it's far from guaranteed. Investors betting on the company are essentially wagering that its technological edge and integrated manufacturing model will allow it to weather the coming period of industry overcapacity and emerge as one of the winners. That's a reasonable bet, but it's definitely still a bet.