Samsung Foundry has started raising the prices of certain semiconductor fabrication services by up to 15%, taking advantage of a demand that is filling the most advanced production lines in the industry. The increases affect 4- and 5-nanometer processes and also some products manufactured at 8 nm. This move comes at a particularly favorable time for Samsung: the available advanced capacity in the market has become scarce, and the company itself acknowledges a growing demand from the United States and China, largely driven by artificial intelligence (AI), high-performance computing, and HBM memory.
The key points of Samsung Foundry’s price hike in 20 seconds
- Samsung is implementing increases of up to 15% on certain new chip manufacturing orders.
- The hikes affect 4 nm, 5 nm nodes, and some 8 nm products.
- Demand for AI is occupying capacity both directly and through the base dies used in HBM memory.
- Samsung recently confirmed revenue growth in Foundry and expects to increase its sales by double digits in the second half of the year.
- The capacity shortage in advanced technology presents an opportunity to close the significant gap still existing between Samsung and TSMC.
The new rates have not been officially announced by Samsung. The information comes from sources familiar with the negotiations, as the company maintains a policy of not commenting on prices agreed with specific clients.
According to these sources, new orders placed in July via the SF4 4 nm process registered monthly increases of between 10% and 15% for clients in China and the United States. In Taiwan, the increases would range between 5% and 10%.
The SF5 5 nm process also reportedly increased between 10% and 15%, while certain wafers produced with 8 nm technology are said to have gone up by nearly 10%.
Beyond the specific figures, this change illustrates how the race for AI is beginning to shift the balance in the global semiconductor manufacturing market.
Advanced capacity saturation benefits Samsung
TSMC continues to dominate with a vast lead in third-party chip manufacturing, especially in the most advanced processes used for CPUs, GPUs, and AI accelerators.
This position has a consequence: a growing portion of its capacity is committed to the enormous demand associated with AI computing.
When clients cannot find enough capacity at the Taiwanese manufacturer, they need to look elsewhere. Samsung Foundry is one of the few companies capable of competing in advanced processes, alongside Intel Foundry.
This does not mean TSMC has suffered a “collapse.” The situation is actually the opposite: an extraordinarily high demand is limiting the availability of certain manufacturing technologies.
Samsung is beginning to benefit from this scenario.
In its Q2 2026 financial results, Samsung confirmed that its foundry business had experienced revenue growth driven by demand for base dies for HBM memory and strong orders from U.S. clients. For the second half of the year, it anticipates double-digit revenue growth supported by increased demand from the U.S. and China.
Samsung also states that it continues to secure new designs from major clients, including HPC projects using its 2 nm technology.
This shift is especially significant for a division that has been trying for years to reduce the huge gap with TSMC.
AI is occupying factories even when not producing GPUs
The pressure of artificial intelligence on semiconductor factories is broader than it might seem.
Not all AI-related capacity is used directly for manufacturing GPUs.
A good example is HBM4, the new generation of high-bandwidth memory aimed at AI accelerators.
These memories stack different layers of DRAM over a logic component called an base die. Samsung is using its 4 nm process to manufacture its HBM4 base die. The company began commercial production of these memories in February 2026.
This creates an interesting situation within Samsung itself.
Its memory division needs capacity from Samsung Foundry to produce components for HBM. At the same time, the foundry must fulfill orders from external companies also seeking capacity in these same processes.
The company forecasts that its HBM sales will multiply by more than three times in 2026 compared to 2025 and is already expanding its HBM4 production capacity.
The growth of AI is thus simultaneously pressuring different parts of the supply chain: DRAM memory, HBM, advanced logic processes, encapsulation, and custom chip fabrication.
For Samsung, there is also an advantage that few competitors can replicate: the company has within its group memory, foundry, and advanced packaging technologies.
Samsung is trying to turn this integration into one of its key selling points for securing new contracts related to AI infrastructure.
From losing money to seizing an opportunity with AI chips
The increased utilization of factories can also help address one of Samsung Foundry’s long-standing issues.
A semiconductor fab requires billions of dollars in investments and huge fixed costs. Operating lines below capacity significantly hampers the ability to recover these investments.
The scenario changes when factories are fully loaded.
Higher utilization allows spreading those costs over more wafers and, if the manufacturer can also raise its prices, profit margins can improve even further.
Samsung has not confirmed that its foundry division will return to profitability by a specific date. Therefore, any forecast setting that date in 2027 should be considered an estimate, not an official target.
However, there are signs of business improvement.
In its Q2 results, Samsung highlighted the expansion of new designs with major clients and projected double-digit growth in Foundry revenue for the second half of 2026.
The company also plans to invest more than 110 trillion won in facilities and R&D during 2026, as part of its strategy to strengthen its position in AI semiconductors.
Samsung seeks something more than just higher prices
The real prize for Samsung is not necessarily that 15% increase.
The shortage gives the Korean manufacturer a chance to gain something much more valuable: to have leading chip designers test their manufacturing processes.
Switching foundries is not just about sending the same design to another factory. Chips must be adapted to the specific technologies, libraries, and features of each process.
Once that work is done and production validated, the client can decide to keep some of its future products with that manufacturer.
Samsung is trying to capitalize on this moment to expand its client portfolio in AI and high-performance computing.
For example, in July, it announced an agreement with Broadcom to expand collaboration in memory, foundry technologies, and advanced packaging for future AI infrastructure. The two companies estimate the collaboration could exceed $200 billion by 2030, though this is based on a memorandum of understanding and not guaranteed revenues.
Samsung is also progressing in 2 nm processes aimed at HPC and preparing new generations for upcoming years.
Older chips are also subject to increases
Another significant point is that the increases are not limited to the most modern nodes.
Certain products manufactured at 8 nm have also reportedly seen increases close to 10%.
These mature processes continue to be used in many chips where employing the most advanced manufacturing technology would be unnecessarily expensive.
This demonstrates that the current pressure on the industry cannot be explained solely by the race to 2 nm.
AI is reshaping investment and production priorities throughout the supply chain. Manufacturers reserve resources for higher-demand and more profitable products while customers compete for capacity across different process generations.
Samsung’s own strategy reflects this situation.
By the second half of 2026, the company plans to increase sales of base dies for HBM, accelerate its second-generation 2 nm process for mobile devices, and continue winning projects related to AI and HPC with advanced nodes.
Additionally, Samsung is preparing for even greater integration between memory and foundry. For HBM5, it has plans to use a base die manufactured using a 2 nm process, compared to the 4 nm technology used in HBM4.
Artificial intelligence is thus enabling Samsung Foundry to recover ground at a time when its main rival is facing demand difficulties that are hard to absorb.
TSMC continues to hold a huge advantage in market share, technology, and customer base. A temporary price increase at Samsung does not change that situation.
But having more occupied factories, raising rates, and winning new designs can help improve the economics of its foundry business. For Samsung, which has been trying for years to convince major chip designers of a viable alternative to TSMC, capacity shortages could end up becoming an opportunity it could hardly have created on its own.
Frequently Asked Questions
How much is Samsung increasing chip manufacturing prices?
Available information points to increases of up to 15% on certain new orders. The exact amount varies by process and region.
Which Samsung Foundry processes are seeing price hikes?
Increases have been noted for 4 nm and 5 nm processes, as well as some wafers produced with 8 nm technology.
How does AI relate to these price increases?
AI is driving higher demand for accelerators, custom chips, and HBM memory. The latter also uses advanced foundry processes to manufacture components like its base dies, further increasing pressure on available capacity.
Can Samsung catch up to TSMC with this situation?
In the short term, a significant gap remains between the two companies. However, the high demand offers Samsung the opportunity to boost factory utilization and acquire new clients in AI and high-performance computing.
