Samsung Electronics has notified customers that it will end all of its 8-inch wafer foundry operations by the end of 2030, according to Korean semiconductor outlet The Elec, which reported the news on November 2. Samsung had previously signaled a gradual reduction of its 8-inch production lines, but this is reportedly the first time a specific end date has been communicated. TSMC has also said it will close some of its 8-inch fabs, but the impact of the two companies' capacity cuts on the market may differ. The capacity Samsung will ultimately shut down is estimated at up to 2.4 million wafers a year, equivalent to about 70% of the 2025 capacity of TSMC affiliate VIS. For TSMC, a production transfer to VIS has been reported, whereas no clear destination for Samsung's output has emerged, which could further tighten the supply of mature-node semiconductors.
Samsung to Shut Down Completely by End of 2030; TSMC Closing Some Fabs
According to The Elec, Samsung Electronics recently told its 8-inch foundry customers that it plans to halt all related production lines by the end of 2030. The company is reportedly withdrawing from the 8-inch business and will concentrate its personnel and capital investment on advanced manufacturing processes using 12-inch wafers. Under the original plan, only one of the three 8-inch fabs in Giheung, South Korea (S6, S7 and S8), namely S7, was to be closed during 2026, while the other two would keep operating. If the latest report is accurate, the downsizing policy has expanded into a complete withdrawal by the end of 2030.
At its first-quarter 2026 earnings call in April, Samsung Electronics Foundry Division Vice President Kang Seok-chae explained that the company would progressively close 8-inch lines used for power semiconductors, display driver ICs (DDIs) and image sensors. The company has also stopped accepting new orders for products using 8-inch processes since 2026. However, the end-of-2030 deadline for complete shutdown is, for now, based on The Elec's reporting from industry sources and should be distinguished from an official announcement by Samsung Electronics.
TSMC, for its part, has officially disclosed plans to close specific fabs. At its earnings call on April 16, 2026, it said it would phase out Fab 2, which handles 6-inch wafers, and Fab 5, an 8-inch fab focused mainly on gallium nitride (GaN) semiconductors. The freed-up facilities are planned to be used to support advanced processes, and the company explained that it will retain sufficient capacity to meet existing customer demand even after both fabs close.
There is also a possibility that TSMC's 8-inch business will see a larger-scale production transfer. According to a report in Taiwan's Economic Daily News on February 10, 2026, TSMC's annual 8-inch wafer capacity is about 5 million wafers, and there is a view that roughly 80% of it will be progressively transferred to affiliate VIS over the next several years. Citing analysis from IC design companies, the paper reported that production would be taken over through order transfers, relocation of manufacturing equipment and technical cooperation. It also reported that TSMC has sold 8-inch manufacturing equipment to VIS twice in recent years. However, the roughly 80% transfer is an industry outlook, not a plan TSMC has officially committed to.
Up to 2.4 Million Wafers a Year of Capacity Could Disappear
Samsung Electronics' 8-inch fabs use mature processes from 65nm to 180nm to manufacture DDIs, power management ICs (PMICs), microcontrollers, image sensors and more. These chips do not necessarily require cutting-edge miniaturization; manufacturing cost, voltage characteristics and compatibility with existing designs matter more. As a result, demand for 8-inch fabs remains even as advanced chip production shifts to 12-inch wafers.
Supply is especially tight for power-management-related semiconductors. According to TrendForce's analysis of May 7, 2026, products such as PMICs and power discretes still depend heavily on 8-inch processes. While power demand from AI servers, general servers and edge AI devices is growing, major foundries are cutting 8-inch capacity, so the average 8-inch utilization rate of the world's top 10 foundries is expected to rise from about 80% in 2025 to nearly 90% in 2026. TrendForce forecasts that global 8-inch capacity will keep declining year on year through the first half of 2027, with utilization staying above 80%.
In this supply-demand environment, Samsung's full withdrawal is of non-trivial scale. According to The Elec's January 2026 reporting, closing S7 in Giheung was expected to reduce the company's 8-inch capacity from about 250,000 wafers per month to under 200,000. Annualizing the 200,000-per-month ceiling gives up to 2.4 million wafers of capacity subject to the eventual full shutdown. Meanwhile, the Economic Daily News reported VIS's 2025 average monthly capacity at 286,000 wafers, equivalent to 3.432 million wafers annualized.
Comparing the two, the up to 2.4 million wafers that Samsung will halt equal about 70% of VIS's 2025 annual capacity.
- Samsung capacity subject to closure: up to 200,000 wafers/month × 12 months = up to 2.4 million wafers/year
- VIS 2025 capacity: 286,000 wafers/month × 12 months = 3.432 million wafers/year
- Ratio: 2.4 million ÷ 3.432 million ≈ 70%
Both figures are based on equipment capacity, not actual shipment volumes. The VIS figure is a reference point for grasping the scale of the capacity reduction, not a comparison of the two companies' market shares.
For TSMC, it has been reported that about 4 million wafers, 80% of its roughly 5 million wafers of annual capacity, could be transferred to VIS over several years. If realized, VIS would take over part of the wafers that would no longer be produced at TSMC's fabs, within the same 8-inch market. However, transferring equipment and orders takes time, and there is no guarantee that the same capacity will be maintained before and after the transfer. Moreover, existing capacity moving to another operator does not in itself mean a net increase in total market supply.
By contrast, no large-scale transfer destination has been identified so far for Samsung Electronics' 8-inch output. According to The Elec, DB HiTek and SK keyfoundry, the leading domestic 8-inch foundries in Korea, have already allocated much of their capacity to existing customers and have limited room to take on additional orders from Samsung. Taiwan's 8-inch fabs also have little spare capacity, and moving production to Chinese foundries could run into restrictions on end customers' use of Chinese-made chips.
Migrating to 12-inch processes is not easy either. Samsung's 8-inch customers have mainly used 90nm to 180nm processes, selecting among them according to manufacturing cost and product characteristics. Moving to a different 12-inch manufacturing process may require design changes and re-verification of performance and reliability, and could raise manufacturing costs. In analog and power semiconductors in particular, specialized processes are used to support high voltages and other requirements, so products cannot be transferred simply by shrinking circuit linewidths.
In response, some fabless chipmakers are reportedly moving to secure Samsung production slots early and build up inventory. However, stockpiling requires additional funds and storage costs and is not a permanent solution. In the longer term, they will need to secure alternative manufacturing lines and re-qualify their products.
Can 70% Utilization and a Four-Year Runway Ease the Shortage?
In assessing the impact of Samsung's withdrawal, it is important to distinguish between capacity and actual output. The Elec reported in January 2026 that utilization at Samsung's 8-inch fabs was about 70%. Applying that figure simply, actual output against 250,000 wafers per month of capacity would be 175,000 wafers a month, or about 2.1 million wafers annualized. Applying the same utilization rate to the 200,000-per-month ceiling after S7's closure gives 1.68 million wafers a year.
Assuming utilization stays at 70%, the output lost through complete withdrawal can therefore be roughly calculated at about 1.68 million to 2.1 million wafers a year. This is a rough estimate using early-2026 utilization and is not a forecast of actual output through the end of 2030. Even so, it is equivalent to about 50% to 60% of VIS's 2025 annual capacity, a volume that other foundries could not necessarily absorb easily.
Another mitigating factor is the roughly four years remaining until full shutdown. An industry source The Elec spoke to suggested that, given Samsung has provided a four-year preparation period, it is hard to call the decision unilateral. At the same time, the same source pointed out that domestic Korean capacity is fully booked, transfers to China are difficult, and fabs in Taiwan and the United States have no spare room. Even with time before the transition, there is no guarantee that customers will find a contract manufacturer with the right process and capacity.
That said, it is not the case that mature-process supply capacity as a whole will fail to grow. According to TrendForce, Chinese foundries are expected to account for about 70% of the expansion of 12-inch mature-process capacity. If product designs and manufacturing conditions are compatible, moving production from 8-inch to 12-inch remains an option. Such migration is already underway for DDIs and image sensors, and Samsung itself is moving production of both products to 12-inch. TSMC has also said it will expand mature-process capacity for CMOS image sensors at JASM's Fab 1 in Japan.
For Japanese electronics makers, procurement prices for PMICs and similar parts deserve particular attention. As AI server power demand has pushed up 8-inch utilization, TrendForce noted in May 2026 that foundries had begun passing price increases on to customers. In addition, TrendForce's June 30 analysis showed 8-inch foundry prices rose an average of 5% to 15% from the first to the second quarter of 2026, and the upward trend could continue through 2027. If Samsung's complete withdrawal proceeds as planned, whether capacity constraints persist beyond 2028 could be one factor shaping price trends.
Still, Samsung's exit will not necessarily lead to a long-term supply shortage. If transfers to alternative foundries and migration to 12-inch processes progress, part of the lost capacity could be offset. The indicator for testing this view is the average 8-inch utilization rate of the world's top 10 foundries. If it falls below 80% from 2028 onward, the view that 8-inch supply-demand tightness will persist would need to be revised. A postponement by Samsung of the end-of-2030 deadline would also change the pace of the expected capacity decline.
TrendForce's current outlook for 8-inch capacity and utilization focuses mainly on the period through the first half of 2027 and does not fully cover Samsung's end-of-2030 shutdown deadline. The long-term market impact of the withdrawal plan therefore cannot be judged without checking utilization rates from 2028 onward and the progress of production transfers. While TSMC is expected to shift existing 8-inch production to an affiliate, Samsung is heading toward full withdrawal with no clear transfer destination reported. How much this difference will affect supply capacity and prices in the 8-inch foundry market is the key question going forward.